mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-01 15:35:12 +01:00
6a4c193906159c96eb9730bfbc209fa07ead3f7b
530
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 DottaandPaperclip
|
e40de0c9e8 |
perf(ci): refresh test weights and narrow graph dependencies (#18464)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Contributors need fast, reliable CI feedback. > - Diff-aware CI selects test files and partitions them into runner jobs. > - Stale cost estimates put several slow suites on the same runner. Broad imports also couple small tests to large modules. > - This PR refreshes test costs, runs independent graph discovery in parallel, and narrows three dependency boundaries. > - Selection safeguards and CODEOWNER-only AWS routing stay in force. ## Linked Issues or Issue Description Refs #18398. This follows the audit of [PR #18427's verification run](https://github.com/paperclipai/paperclip/actions/runs/41169978784). The audit found 403 selected general-server files without timing weights. Three suites measured at 232s, 122s and 116s shared one shard because each received a roughly 2s fallback weight. Local graph discovery across four projects took 43.7s serially and 21.7s with two workers. A reverse-order repeat took 37.7s serially and 22.6s with two workers. Both comparisons produced identical selection results. This is a local measurement, not a claim about production CI speed. ## What Changed - Refresh general-server costs for 1,285 of 1,309 current files and serialized costs for all 163 files from two successful protected-master Cloud readiness runs. Retain prior costs for unmeasured files, including dedicated chat and native-runner suites. - Record passing server file costs, including collection, setup and environment preparation. A weekly hosted workflow validates trusted source runs and opens a timing-only PR. It tests shard coverage, pauses while a refresh PR is open, and never merges automatically. - Discover at most two project graphs concurrently in isolated, heap-limited Node processes. Cache repeated inventory reads within one selection. Worker errors and timeouts retain the full task plan. - Fetch only the merge commit and parents for the selector. A real shallow-clone regression proves the complete diff still includes more than 300 files, renames and deletions. - Share the identical log viewer and run details between both AgentDetail pages. The viewer test imports the shared implementation and runs each of its 13 distinct cases once. - Extract the pure API error class while preserving the API client's public exports. Import delivery work signals through their existing narrow DB export. ## Verification - PASS: `pnpm -r typecheck` and `pnpm build`. - PASS: 284 CI selection, timing, shallow-diff, sharding and routing tests. - PASS: 74 further cache, routing and graph-proof tests. - PASS: 94 targeted UI assertions across the viewer, run invocation, API errors/client and progress helpers. All 76 assertions in the 12 server suites made independently selectable also pass. - PASS: `pnpm check:token-gates`, module-boundary checks, YAML parsing and `git diff --check`. - PASS: [Full PR CI and required `ci / verify` aggregate](https://github.com/paperclipai/paperclip/actions/runs/41206683345), including all 12 general-server shards, all nine serialized shards, workspace groups, chat suites, Runner lanes and build checks. - PASS: [GitHub hosted real-workspace selector and cache/routing qualification](https://github.com/paperclipai/paperclip/actions/runs/41206683205). - PASS: Both Greptile and Paperclip reviews are 5/5 with no actionable findings or inline comments on `7dc65c51b8`. The PR has no merge conflicts. - PASS: The production refresh parser accepted all 107 structured file-cost records from a real successful server CI job, including setup/loading cost. - Local `pnpm test:run` completed all six chat shards successfully. The duplicate monolithic local run was stopped after the complete remote suite passed; it was not completed locally. The focused log-viewer command improved from 18.84s (26 duplicate variant cases) to 8.49s (all 13 distinct behavior cases). The dependency audit confirms 12 server files and the API-error test no longer need conservative always-run retention. Repartitioning the audited PR with the new costs changes the slowest shard from 810s to 591s of summed historical test execution (1,124 of 1,142 files measured). This is a workload simulation, not measured new CI wall time. ## Risks - Two graph processes use more peak memory. Each has a 2 GiB heap limit and a three-minute timeout. Any failure preserves the original task plan. Real workspace qualification runs in GitHub CI. - Historical logs contain test/hook costs only. New reporter records add loading/setup costs after merge. The manifests identify measurement coverage and source runs. These weights change placement, never permission to skip tests. - Runner load adds timing noise. The refresh uses medians and requires at least 90% measurement coverage in each lane. - GitHub may require approval to run checks on a PR created with `GITHUB_TOKEN`. Refreshes remain reviewable and do not merge themselves. - The shared viewer is a mechanical extraction. Both pages retain their existing exports, and the behavior assertions remain covered. AWS runner authorization is unchanged. ## Model Used OpenAI Codex, GPT-6, with repository inspection, reasoning, code execution and GitHub tools. The exact deployment ID and context-window size were not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub references) - [x] My branch name describes the change and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (focused checks and six chat shards; complete suite passed in CI as detailed above) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
d5766e4074 |
perf(chat): scope task questions and batch inbox queue checks (#18422)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Standing chats show recent messages and current task questions. > - Recent-message pagination bounds history work, but it does not bound question work. > - The question read built the full company attention feed before it selected the chat's tasks. > - Each feed read also checked seeded queue membership with one insert per item. > - This pull request scopes question collection in the attention owner and batches queue membership checks. > - Chats keep their current question cards and initial layout while doing less work. ## Linked Issues or Issue Description Follow-up to #18358 and #18323. Related work: #18312. **What happened?** A hosted standing chat remained slow after recent-message pagination shipped. On commit `4eaa435bf4`, an embedded-browser reload made the conversation usable at about 8.6 seconds. Recent comments finished at about 1.9 seconds. The pending-question request held the initial reveal until about 8.4 seconds. That request called the complete company attention feed and then filtered its results. The separate badge summary also did broad work. **Expected behavior** Chat question cost should depend on the relevant tasks. Reading an existing queue must not require a separate insert for every member. Current access and decision policy must remain in force. **Steps to reproduce** 1. Create a standing chat with questions on tasks assigned to its agent. 2. Add many unrelated tasks and pending questions to the company. 3. Open the chat from a fresh application cache. 4. Compare recent-comment arrival with the pending-question response and content reveal. **Paperclip version or commit** Baseline: `909e29dd071b1aeeb24c2933af049da90a6066d3`. Its affected services match the observed hosted revision `4eaa435bf4`. ## What Changed - Add an internal task-question projection to the attention service. Filter pending interactions by company, task IDs and question kind before card enrichment. - Keep the same access, audience, dismissal, queue, snooze, retention, archive and training-example logic. Ranks are local to the returned question list. - Use the projection from the conversation receipt service. Keep current chat audience and session checks there. - Batch seeded membership inserts in groups of 500. Keep the queue lock, conflict protection, audit events and activity logging for each new membership. Existing members produce no new audit records. - Add real PostgreSQL regression cases and a reproducible paired benchmark. No schema migration or UI reveal change is needed. ## Verification - `pnpm -r typecheck` passed. The final server typecheck also passed after the queue change. - `pnpm build` passed. - Real PostgreSQL service coverage: 57 attention and chat-handoff tests passed. Seven queue-route tests passed, including concurrent reads, mixed existing/new items, duplicates, disabled seeds and audit preservation. - Red/green proof: the unrelated-attention regression fails with the baseline conversation implementation and passes with the scoped read. - `node scripts/check-module-boundaries.mjs` and `git diff --check` passed. - Full `pnpm test:run` was attempted and stopped with exit 130 after embedded PostgreSQL startup failures and 17 failures in the untouched native-session executor suite. It is not a local pass. The focused Docker/PostgreSQL results above completed before Docker Desktop became unavailable. A direct native PostgreSQL probe identified the host semaphore limit (`could not create semaphores: No space left on device`). No shared kernel settings or other test clusters were changed. [Full CI](https://github.com/paperclipai/paperclip/actions/runs/41162314995) passed on `2809f096b3fb2c01b44a96a1b82a27809c9359df`, including all eight browser shards. Final status: 55 successful checks/statuses, three skipped and one neutral. Greptile is 5/5 with no findings or unresolved threads. The fresh local browser retest is blocked because Docker Desktop cannot start and native PostgreSQL cannot allocate semaphores. The original hosted reproduction above is complete; no candidate browser pass is claimed. Local service benchmark: PostgreSQL 17.11, two relevant tasks, five alternating pairs per scenario, after warmup. These are local service timings, not a claim about deployed latency. | Fixture | Read | Baseline median (range) | Candidate median (range) | SQL calls | | --- | --- | --- | --- | --- | | 1,000 unrelated tasks/questions | Chat questions | 670 ms (632–1,905) | 12 ms (11–15) | 1,076 → 22 | | 1,000 unrelated tasks/questions | Company badge summary | 641 ms (622–717) | 167 ms (161–171) | 1,068 → 69 | | Small company | Chat questions | 38 ms (37–52) | 13 ms (12–16) | 52 → 22 | | Small company | Company badge summary | 34 ms (33–39) | 34 ms (32–40) | 44 → 43 | All question fields and ordering matched, except rank is relative to the scoped list. Summary fields matched except the generation timestamp. The interaction plan changed from 1,002 selected rows to two rows through the existing company/issue/status index. The benchmark loads baseline attention, conversation and queue modules together so the queue change cannot contaminate the baseline. ```sh pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-chat-questions.ts \ --baseline-ref=909e29dd07 --unrelated=1000 ``` ## Risks - Scoped and complete feeds must share policy. This change reuses the same attention collection and finalization path and tests dismissals, named audiences, snoozing, archives and private access. - Batched inserts must audit only new members. The concurrent PostgreSQL test checks unique membership and one event/activity record per new item across batch boundaries. - Full inbox summaries still collect the company feed. The small-company summary timing is unchanged. Other chat requests and transcript hydration can still contribute to hosted latency. - The first reveal still waits for question cards to preserve layout. This follow-up has not been deployed, so its hosted effect remains unverified. ## Model Used OpenAI Codex, based on GPT-6, with reasoning, repository tools, code execution and browser profiling. The exact serving model ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
492d537340 |
fix(native): recover controllers across webserver replacement (#18343)
## Thinking Path > - Paperclip manages AI agents and their work. > - Native execution runs in a remote Runner while the webserver owns its controller. > - Encrypted PRP checkpoints and session snapshots already survive a webserver restart in Postgres. > - Ordinary warm and per-turn admission still needs local markers and supervisor metadata to find that state. > - Replacing the webserver with an empty native controller directory can therefore lose a usable retained session. > - This change adds bounded recovery metadata to the existing snapshot transaction and reconstructs local projections from the selected durable source. > - The benefit is native execution continuity across webserver replacement, with the existing ownership and agent lifecycle policies. ## Linked Issues or Issue Description **What existing behavior does this improve?** Native remote controller recovery and retained-session continuation during a webserver upgrade. **Subsystem affected** server/ — native runtime, checkpoint persistence, recovery scheduling and internal upgrade reporting. Product E2E qualification also changes. **Current behavior** Without both the local PRP marker and supervisor file, normal continuation can return before discovering the saved Postgres checkpoint. The snapshot already exists in Postgres, but its local reuse compatibility envelope cannot be reconstructed. **Proposed behavior** Select the existing durable source, validate its identity and ownership, restore its PRP state and supervisor projection, then use normal authenticated admission. Missing or inconsistent evidence holds recovery. It does not authorize fresh provider work. **Reason and benefit** A replacement webserver must recover active execution and later continue its retained session without the old server's native files. This is one component of deployment continuity. HTTP traffic switching, workspace storage migration and Railway rollout remain separate. **Breaking changes** No new public endpoint, schema migration or database pool. The private run-profile envelope is additive. Disk-independent rollback also requires a reader for the staged controller format and compact checkpoint part; the first production release containing these readers is the floor. Older file-backed sessions need an owner checkpoint or the bounded internal importer. **Additional context** Related foundations: #17401, #17916, #17254, #14621, #14622. The implementation contract and open acceptance gates are in [the dated plan](doc/plans/2026-10-30-native-controller-checkpoints.md). ## What Changed - Persist a versioned recovery envelope, at most 4 KiB, beside the redacted session snapshot in its existing transaction. Keep it out of public run projections. - Compare frozen admission separately from prepared attachment instructions during staged recovery, preserving exact template/identity/authority checks. - Bound the existing lazy checkpoint pool at three connections after measured two-slot burst backlog; keep request/renewal capacities and SQL operation counts unchanged. The tuning is a separate commit. - Discover and restore the exact selected retained source without local native markers. Preserve the 32-candidate window, progress barriers and current authority checks. - Fence snapshot and result writes. Acquire coordinator, issue and run locks in order and check expiry after lock waits. Combine snapshot ownership/read work into one statement. - Reuse existing recovery deadlines and drain/resume scheduling. Retained-owner waits preserve the original unstarted run and actual lease deadline. Stop, ownership and provider-progress fences precede deferral. Add bounded native evidence reporting, retained-source import and retention protection. - Admit a settled warm source with an authenticated idle goal-read suffix while retaining active-work and mixed-authority holds. - Permit lifecycle `turn.stop` checkpoints after accepted results under existing ownership fences. Claim the existing successor coordinator before retiring the old warm session, and stage shared attachments only after verified retirement. Preserve omitted isolated-home defaults and existing fresh skill-selection order during recovery without hiding real configuration changes. - Carry the exact durable six-field PRP binding during active recovery after a healthy warm handoff. Keep the Runner mismatch guard and preserve opaque attachment IDs. - Preserve an already verified active or idle warm process during remote preparation; revalidate its proof/liveness and retain later authenticated PRP attachment. Ordinary resumed leases retain suspension checks. - Drain retained source events under their original authority, reconcile immutable source accounting and block unknown source tool dispatch before target work. - Persist compact staged handoff intent through authenticated adoption and recover direct/listener command-loss windows without another Runner or duplicate attachment. - Rebuild suspended lifecycle receipts for resumed prepared tasks, and retry missing lifecycle reads after partial PRP materialization without overwriting contradictory evidence. - Add normal warm/per-turn diskless continuation coverage, real Postgres races and SQL-cost checks. Add a reproducible paired load workload. - Give only the explicit outage fixture a 15-minute provider idle window, with a red/green catalog regression preserving all production timing and lifecycle policies. - Publish the goal-less task-session pointer in the existing post-acceptance terminal checkpoint transaction, with current ownership, accepted-result, reset and newer-progress fences. Keep a private bound receipt so late completion cannot duplicate or resurrect publication. - Use a retained warm sandbox in the dedicated continuity fixture, and accept an exact output token followed by an explanation. Keep the strict identity oracle. - Extend the explicit Daytona cell to a five-minute active outage and another replacement after completion. Require retained provider identity, a new tool receipt and a useful follow-up result. Add an owned loopback Postgres E2E fixture for hosts that cannot start embedded Postgres. ## Verification **Merge recommendation — `fc68584ce6231a02e47ecd743f55b036ccea034f`: ready for the qualified native-controller component.** Full CI, fresh review, local application paths, controlled performance qualification and the strict live Linux/Daytona case pass. Independent resource cleanup is complete. Broader rollout, browser reconnection and the unexecuted transport/platform combinations remain integration gates. Integrated 484 passed full local recursive typecheck/build and ordinary active→completion→cache-loss→C component qualification. Fresh review then found a valid attached-wake failure: preparation appends attachment instructions before the verifier compares execution to the frozen admission. 9f4 fixes that by separately checking the pre-preparation admission and exact prepared execution. Both hashes are captured privately per admission; the full template, identity, configuration and authority checks remain. Its regression fails before the fix; 36 focused cases, 42 staged source/target cases and server typecheck pass afterward. The review thread is resolved. Exact-head full CI passes ([run 41168228822](https://github.com/paperclipai/paperclip/actions/runs/41168228822): 48 successful jobs, one intended skip); the complete `pnpm test:run` partitions are forced, with the five native suites in the Runner lane. This is not a claim that an unsharded local command passed. The fresh exact-head Greptile check succeeds with 5/5 and all six threads resolved. The original two-connection database comparison closes **steady PASS and burst FAIL**. Every relative CPU/WAL/request-p99 gate passes and all operations/ownership remain correct, but candidate arm 3 drains in 1,015.99 ms and baseline arm 4 in 1,059.67 ms against the unchanged 1,000 ms absolute gate. Those results remain failed. A separate 2→3→4→2 diagnostic reproduces drain times 1,051.57 / 956.79 / 882.02 / 1,065.72 ms: three is the smallest tested passing cap, and the repeated two-cap failure survives run-order/cache effects. Separately reviewable `fc68584ce6` changes the existing lazy checkpoint pool default to three connections. Request capacity stays ten and liveness two; idle release, shutdown fencing, transaction counts and pool count are unchanged. Thirteen DB lifecycle/cap tests and DB typecheck pass. Full recursive typecheck/build pass at exact |
||
|
DottaandPaperclip
|
59da15066a |
feat(runner): deliver native Kimi Code (#18391)
Deliver native Kimi Code setup, exact models, scoped credentials, verified packages and runtime packs. Preserve current upstream release attestations and restore verified same-run remote retry dispatch without weakening process ownership. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
100b364264 |
Preserve selected Google services when setup resumes (#18048)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Connections control which external services an agent can use. > - Google Workspace setup lets a person choose services and read or write access. > - A draft lost that choice after a declined or expired OAuth handoff. > - Finish setup could then show all services with write access instead of the saved narrow choice. > - This pull request saves the exact draft choice without treating it as granted access. > - The benefit is safe setup recovery that preserves the person's choice and existing account limits. ## Linked Issues or Issue Description Refs #15545. The duplicate search found no open PR for Workspace draft consent recovery. Related #18000 concerns satisfied gateway access requests, not saved OAuth setup intent. **What happened?** A cold Workspace draft resumed with the new-account service defaults. The selected request lived only in short-lived OAuth state. Denial consumed that state. The picker ignored draft connections. **Expected behavior** Finish setup restores the same person's saved service choice. It does not use another person's choice, a broad profile hint, or a credential grant as draft intent. An old draft with no reliable choice asks the person to choose services. **Steps to reproduce** 1. Start a new Google Workspace connection. Select Gmail read-only. Turn off Drive and Calendar. 2. Continue to Google, then decline authorization. 3. Close the setup page. Return through Finish setup. 4. Open the service picker. Before this fix, all three services can be enabled with write access. **Paperclip version or commit** Reproduced in tests on master `820c458465fa40137c81fe510f3e6dc573b4f895`. **Deployment mode** Authenticated and local source builds. The source regression uses real PostgreSQL fixtures and the production React setup components. Live provider acceptance of this repair is still pending deployment. ## What Changed - Save validated, non-secret draft intent before managed or customer-owned OAuth. Bind it to company, connection, actor, grant kind, and subject. - Restore that intent after denial, state expiry, and cold Finish setup. Keep active reconnect bound to its saved grant. Do not change new-account defaults. - Preserve agent-to-human setup handoffs. Consume only the addressed person's reviewed proposal. Keep other people's proposals. - Lock and merge the current draft before saving intent. Return 409 if its credential identity or method changed during setup. - Preserve the latest saved choice across overlapping customer-owned starts. Merge both current configuration records under short locks. Reject stale client-binding changes. Keep each OAuth state's own requested scopes. - Add shared, service, handoff, picker, and cold-page regression tests. Update the connector contract and existing feature plan. - Route the default ungrouped test/release paths through six sequential exact-cover chat partitions. Preserve remaining server/native coverage, explicit grouped CI, selection-first execution and shared-database paths. - Correct the retention test driver so pending owned real I/O does not advance logical deadlines. Pump zero-duration connection timers and add a held-I/O regression; keep production scheduling and original timing assertions unchanged. The shared helper is needed by the server and picker to validate the same narrow binding. It does not resolve credentials or grant access. No production schema migration, new route, provider scope or deployment configuration is added. The test-harness changes do not change production timers. ## Verification ### Current candidate: source qualification passed; live acceptance pending - Current published candidate: `82d13d4d698a5579378b5d506cb7a1f9cb64b5c8`, tree `78d6a821ce53f481fe3e9991edce56f0acffb22f`. It combines the Workspace repair with launcher commit `bf29c9` and retention-test commit `e25987`, then integrates master `493e229d93ba7798e80152f4549e2e9177ce50ae`. The integration brings 40 upstream commits across 713 files. One launcher conflict keeps upstream validated CI selection and its selection-first path, then applies six default chat partitions only without a selection. The authored PR diff has 16 files, 833 additions and 100 deletions; it adds no authored lockfile or `.github` changes. The public head is confirmed at this revision. Fresh CI and review passed as recorded below. Source qualification is complete using exact-tree full CI and the local results below. The separate local whole command was interrupted by host disk protection and is not called a pass. Live repaired-draft acceptance remains pending deployment. - Exact clean `82d13d4` passed one Node 24.19.0 builtin-only run: 41/41 cases (37 launcher and four CI-workflow cases), zero failures/skips. HEAD, tree and relevant file hashes stayed unchanged and clean before, at start and after. This proves launcher fixtures, not real chat execution, current typecheck/build, the dependency-backed selection graph, or full feature acceptance. - The new default ungrouped whole command uses six sequential exact-cover chat partitions, then the remaining server suites, workspaces and serialized suites: 28 expected invocations. It preserves all native suites, including upstream usage coverage. The plain ungrouped release launch path also changes. Explicit grouped CI, validated selection and shared-database execution stay unchanged. Launcher hard RED had three expected body failures and two unchanged passes; the repaired focused five and final 37 passed. A later 41-case staged-merge pass is retained separately; the clean 82 run above is the frozen proof. - Fresh [CI run 41153113944](https://github.com/paperclipai/paperclip/actions/runs/41153113944) completed successfully on October 31 at 01:14:53 UTC. Current head has 52 successful check runs, four conditional skips and no pending/failing checks; legacy Snyk also passed. The skips are test-file selection, the optional paired benchmark and two Storybook checks. The full server/chat/workspace/serialized matrix, all eight e2e shards, normal build, build-gap typecheck, Runner checks and final aggregate gates passed. Build and all three chat job logs identify checkout `e5e831fd3a8534d0d0f1e892118c0f8fbe12dd92`, merging this head into `493e229d93`. Its Git tree is exactly `78d6a821ce53f481fe3e9991edce56f0acffb22f`, identical to the clean candidate. The six real chat partitions passed 314, 358, 273, 239, 321 and 311 cases: 1,816 total. Other cases shown as skipped in each partition were selection exclusions, not extra passed cases. This is current tree-equivalent CI proof, not a pass of the separate default ungrouped local whole command. - Fresh [Greptile summary](https://github.com/paperclipai/paperclip/pull/18048#issuecomment-6473856992), edited October 31 at 01:05:04 UTC, explicitly reviews exact head `82d13d4` at 5/5 with no actionable code issue. Its exact-head check passed. The prior overlapping-start discussion is resolved. This review is source-only, not provider or live acceptance. - Exact clean `82d13d4` passed full recursive typecheck and normal `pnpm build` on the independent darwin/arm64 snapshot. Typecheck finished on October 31 at 01:27:33 UTC; build finished at 01:28:33 UTC. Both exited 0. Normal native preparation and code signing passed. Source HEAD, tree and tracked/untracked state stayed unchanged and clean. `pnpm check:token-gates` also passed all token, palette and ClipLab checks on this candidate. The earlier forced-clone `ENOSYS` preflight remains recorded as setup failure, not an assertion result. - The unchanged default `pnpm test:run` started at 01:51:33.796 UTC on clean exact `82d13d4`. All six real chat partitions passed: 358, 314, 273, 239, 311 and 321 cases, or 1,816 actual cases. The original FIFO and ambiguous-retry cases both passed under their unchanged 15-second limits. General-server then reported 37 completed file rows with 5,438 passed cases and five default remote-continuity skips, including all 498 tool-access service cases and all 83 connection-route cases. At 02:39:44 UTC, the unchanged 5-GiB host-disk guard terminated the owned run. The child ended with SIGTERM and the container/harness exited 143, with no OOM or recorded assertion failure. Only seven of 28 RUN invocations started. Retention had no completed local row. Later invocations did not run. This whole local command is interrupted/incomplete, not passed. - A coverage check found no default suite missing from current full PR CI. The maintained launcher tests prove exact coverage for all 12 PR server shards, chat and five native suites; all configured workspace projects; and all nine serialized shards. The final Runner CI shard executes the five native server suites. The 41 local launcher/workflow cases cover six sequential chat children, per-child isolation, native retention, fail-fast propagation and unchanged explicit selection/shared-database paths. The actual six-chat run above exercises that changed default path. The remaining unqualified local difference is uninterrupted sequential whole-command composition and resource accumulation, not omitted test selection. Current exact-tree full CI supplies broad source-test qualification; no local whole-pass claim or provider acceptance is inferred. - A task-owned remote runner was considered to complete that separate composition check. Its provider rejected the requested allocation with per-sandbox limits of 8 GiB RAM and 10 GiB disk. No sandbox was admitted and no remote tests ran. No tests, timeouts, assertions, database synchronization, source or provider settings were changed to avoid the capacity limit. Historical a0 failures remain below. No deployment or live repaired-draft acceptance has run. ### Bounded follow-up on the prior source - On a0, one finite phase-traced chat diagnostic executed exactly the two failed cases. Both passed (1,374 and 1,767 ms); 1,814 other cases were filtered, not passed. This did not reproduce a blocked await or establish the original monolithic timeout cause. No assertion or 15-second limit changed. - The retention diagnosis preserved an initial zero-body selector failure, then corrected only the selector: one original case passed in 783 ms, with 42 filtered. Trace showed the old settling helper advancing logical time while real I/O was active. A new held-I/O regression then failed hard because the old helper advanced 1,000 ms. This proves a test-driver defect, not the precise cause of the old 5,340 ms full-run gap. - The first manual-clock repair failed the complete 44-case file: 43 passed, one body hit 15 seconds and cleanup hit 30 seconds. Its planned mutant did not run. Source then identified a fresh PostgreSQL connection's zero-duration timer. The corrected driver pumps zero-duration timers while initial readiness or tracked I/O is pending, without advancing logical Date. That pre-integration file passed 44/44 with no skips. A separate scratch-only 5,000 ms normal-continuation mutant failed the original hard `5000 < 3000` assertion; its expected failure is not a test pass. The original page-count, fairness, expiry and quiet assertions stayed intact. The authored repair does not change the production worker. The merged master changes an older retention case and dependencies; these pre-integration results are not a complete 82 retention or whole-suite pass. ### Historical qualification: SHA-bound, not 82 acceptance - Historical a0 candidate: `a0f9089fc1ff466bf4e1eeddc5b1a910d09e380b`, tree `b9d73713319dcd52b9eaa6053008b5190e43fdf1`. It adds the overlapping-start repair to parent `6c18`, which already merged pinned master `1511c92965325249ed6b52dd5be52eb6d736fbfa`. That candidate changed 12 files. Its CI and review passed; its local full qualification remained failed/incomplete. - Greptile's previous `6c18` review found P1 [Latest service choice gets lost](https://github.com/paperclipai/paperclip/pull/18048#discussion_r4425965828), confidence 4/5. Two hard-assert regressions reproduced it on untouched `6c18`: the already-bound path overwrote the later choice, and real first-use client stamping also restored the old choice. Existing OAuth states/grants were not widened, but a later setup could restore an older request. The new repair covers both write boundaries without holding locks over external work or copying another attempt's scopes. - The complete affected service file passed 498 tests with no failures or skips. All 31 required controls ran, including both unchanged RED cases, 14 new client-drift cases, seven earlier identity guards, reconnect/cold-resume cases and upstream transaction tests. The tested three-file patch is byte-identical to this commit; the connector contract update is separate. Independent scoped review found no additional actionable issue. Fresh Greptile reviewed exact `a0f` at 5/5 with no actionable findings. The previous P1 thread is resolved. - On exact `a0f`, the five shared/UI files passed 312 tests. An initial complete connection-management run then failed: 74 passed, eight timed out at the unchanged 15-second limit, and one AI-account verification returned 422 instead of 201. The Google addressed-user handoff passed. The container exited 1 without OOM and stopped before chat/token/whole tests. This material failure remains retained and its cause unclassified; the later controlled recovery below is not a causal explanation. - One diagnostic executed the first failed gateway case and the two AI cases with temporary phase-only tracing. All three passed at 8.516, 5.511 and 3.523 seconds; 80 other cases were filtered. Original hard assertions and timeouts stayed unchanged. No timeout or suspected late provider cleanup reproduced, so the complete-file failure remains unresolved. Vitest exited 0; the harness exited 1 because its trace parser missed ANSI prefixes. All 203 events were recovered from retained logs without rerunning tests. Instrumentation was removed. This is not acceptance. - Historical a0 CI [41097596364](https://github.com/paperclipai/paperclip/actions/runs/41097596364) completed successfully: 52 successful checks, two expected Storybook skips and no failed/pending checks. Legacy Snyk passed. All 48 CI jobs passed, including all 1,816 chat cases and explicit passing lines for the original 20 local failures. [Server 11/12](https://github.com/paperclipai/paperclip/actions/runs/41097596364/job/124207469276) explicitly passed the complete 83-case connection file in 129.49 seconds, including all nine local failure titles. The actual checkout was synthetic merge `06851da6eb969d63e1c7dc9f7c95f08d09fa4ff1`, combining `a0f` with `aa58fa107389e8b9ae5949f7ad815943dfd94292`. This is not standalone local `a0f` qualification and does not erase the nine local route failures. - Linux admission uses Node 24.22.2, pnpm 9.15.4 and Rust 1.97.1. Frozen offline install, exact native hashes and real migrated PostgreSQL 18.1 passed. Native/protocol/build/lock inputs are identical between `6c18` and `a0f`; validated cache reuse retains the actual `6c18` native-build provenance. The selected packaged release binary is not called a fresh `a0f` build. Database synchronization settings stayed on. This admission alone does not explain the route failures. - A fresh Mac snapshot of exact `a0f` passed frozen install, workspace prerequisites, full sequential recursive typecheck and normal `pnpm build`, both on their first attempt. Typecheck completed at 17:43:05 UTC, build at 17:44:43 UTC; both exited 0 with clean tracked/untracked source. Pinned Node 24.19, pnpm 9.15.4, Rust 1.97.1, package concurrency one, Cargo jobs one and a 12 GiB V8 heap ceiling were retained. Largest-process maximum RSS was 7,668,154,368 bytes for typecheck and 10,038,837,248 bytes for build, not aggregate tree peaks. The first build correctly withheld its freshness stamp because 125 ignored, pinned Copilot assets appeared during preparation; exact path/hash analysis explained this. The normal build then recorded a genuine stamp and reused Cargo outputs in 1.18 seconds. No manual stamp, source workaround or gate retry. This is darwin/arm64 static/build proof, not Linux packaging/runtime or full tests. - The historical a0 candidate's token checker passed all four gates across 1,410 files, run directly with pinned Node 24 against the script behind `pnpm check:token-gates`. - The single controlled complete-file comparison passed: baseline `6c18` 83/83 in 287.07 seconds, then candidate `a0f` 83/83 in 293.03 seconds, with zero skips. All nine earlier failure titles passed on both. Both child/container exits were 0, no OOM/interruption; source and native/resolver/PostgreSQL guards matched. This ran only after all owned Mac compilation ended, with fresh processes/databases and identical four-CPU-quota/6-GiB/overlay inputs. No filtering, instrumentation, assertion/timeout/source changes or retry loop. This establishes standalone complete-file recovery, not the cause of the earlier eight timeouts/AI 422. - The first unchanged full `pnpm test:run` on exact clean `a0f` ran from October 30 at 18:01:43 UTC to 21:06:19 UTC. It failed naturally with test and container exit 1, without OOM. General-server reported 1,257 passed, one failed and seven skipped files; 28,633 passed, one failed and 103 skipped tests. Only the first of 22 invocations ran. The later four groups and 21 invocations did not run. The sole failure was the unchanged chat-shard CLI meta-test: its nested projectless pnpm lookup hit the offline network guard. This was a failed full gate, not Google provider acceptance. - A finite environment pair reproduced that one meta-test failure, then passed the same unchanged case after supported `corepack prepare pnpm@9.15.4 --activate` as the existing non-root node user. The cached pnpm payload was identical. Only the missing node-user last-known-good record changed. Core source, dependencies, lockfile, fixtures, assertions, timers and offline guards stayed unchanged. This isolates the nested collection defect; it does not explain the earlier route failures or complete the full gate. - The corrected unchanged full command ran from October 30 at 21:29:03.500 UTC to October 31 at 00:02:44.207 UTC on the same exact clean `a0f`. It failed naturally with exit 1, signal null and no OOM. Source stayed clean before, at start and after. It retained two visible CPUs/quota, 6 GiB, private IPC, normal overlay, matching native inputs, synchronized real PostgreSQL 18.1, and the default five groups/22 invocations. Only the first general-server invocation ran; the other 21 did not run. General-server reported 1,265 files: 1,256 passed, two failed and seven skipped. It reported 28,737 tests: 28,631 passed, three failed and 103 skipped. The two chat cases emitted `Test timed out in 15000ms` at `chat-channels.integration.test.ts:77310` and `:81790`, with durations of 15,045 and 15,058 ms. These were the durable pre-run FIFO notice and ambiguous GitHub text restart/retry cases. The retention case emitted `expected 5340 to be less than 3000` at `skill-version-retention.test.ts:593`. Its real duration was 1,727 ms; 5,340 ms was the recorded fake-time directory-page gap, not a deadline timeout. The timeout stacks do not identify the blocked internal await. The cause of these failures remains unclassified. Inside this same failed run, the complete 83-case connection file and 498-case service file passed. The former Corepack meta-case also passed in 6,396 ms. These component passes do not waive the failed/incomplete full gate. No rerun, cancellation, timeout/assertion change or full-pass claim. A later fresh two-case phase trace passed both chat cases; it did not reproduce or establish the original timeout cause. At that checkpoint, the PR stayed draft. - Prior `6c18` targeted checks passed 930 tests across eight files and token gates. Its selected queue group passed all 53 cases, including the original 20 failures and 12 new named controls. Backlog cases took 10.49 and 8.45 seconds under the unchanged 15-second limit. Another 1,763 chat cases were intentionally filtered. This is historical targeted recovery, not a full pass or proof of the new repair. No timeout or assertion was relaxed. - Prior `6c18` full local qualification failed during serial typecheck: server TypeScript was killed with code 137. Docker confirmed OOM in the 6 GiB container; full tests and build never started. The failure is retained. The new Mac resource plan avoids repeating that undersized run and does not change the shared Docker VM. - Prior CI [41089793558](https://github.com/paperclipai/paperclip/actions/runs/41089793558) completed successfully for synthetic merge `a73bffcfa3ede051961bcabcfa576c703d10c932` (`6c18` plus `abc270ee4c03e13b4d6eaad156a41b4ebed4bb0a`). All 48 CI jobs passed; all 1,816 chat cases passed, including explicit log lines for the original 20. Check runs were 51 success, two expected skips and one Greptile failure; legacy Snyk passed. These results do not qualify `a0f` or erase its local route failures. - Red proof on the unmodified master baseline: cold picker and Finish setup restore broad defaults; managed and customer-owned drafts lack saved intent. The stale setup tests also fail before the lock guard. - Live staging baseline `b25f3e2af8e77d882cd22d36ca55640e6468e46f` also reproduces the ordinary failure: select Gmail read-only only, cancel at Google, reload the connector catalog, then Finish setup. All three groups return enabled at read/write. Authorization was declined; no new Google access was granted. This is baseline failure evidence, not a deployed-fix pass. - Historical affected suites: 800 tests passed before the final lock guard. The final focused service and handoff cases passed 126 tests after that guard. These are not results for the integrated candidate. - Historical `pnpm -r typecheck` and final lock-guard server typecheck passed. The historical a0 candidate passed the full recursive gate separately, as recorded above. - `pnpm build`: passed again on the final lock-guard source in commit `1c31cde0ed199fbf4e41b7649590554f7886dfc2`. - Historical `pnpm check:token-gates` passed; the historical a0 candidate's direct checker pass is recorded above. - `pnpm test:run` on macOS failed on the frozen commit. It exited 1 after 9,432.41 seconds. General server results: 1,195 files passed, 10 failed, eight skipped; 26,706 tests passed, one timed out, 157 skipped. Nine setup hooks timed out while starting embedded PostgreSQL. The retained log records host-wide semaphore exhaustion (`semget`), not disk exhaustion. One unchanged publication-text performance case also timed out. Its cause is not established. Later workspace and serialized groups did not run. This is a failed and incomplete local gate. - The independent full-source Linux run also failed and was incomplete on the same frozen commit. Native and real-database admission passed. Chat integration reported 20 failures in 1,593 tests. Tool-access passed all 479 tests. Native-session-executor passed all 965 tests. Docker interrupted the run after about 48 minutes on October 29. The container exited 143 and was not OOM-killed. The run produced no final group totals or detailed assertion summary. Shell trap metadata reports zero, but this is not a valid test-pass result. Source and evidence remain unchanged. The focused diagnosis below retains those failures; their cause is not assumed from the interruption. Neither historical local full run passed. - A fresh focused replay on the frozen source then executed exactly those 20 case bodies. All 20 failed; this was not a setup-hook failure. The two backlog cases reached the unchanged 15-second timeout. The 18 FIFO cases failed on missing notices or receipt rows. All 18 FIFO cases passed when run alone in a fresh process and database. This supports an order interaction, not a complete cause or baseline conclusion. A separate two-backlog control used RAM-backed temporary storage and real PostgreSQL with synchronization settings on. One backlog case passed in 10.65 seconds; the other still timed out at 15 seconds. No source, assertion or timeout was changed. The full local gate remains open. - Earlier mixed-revision run: it began before the final guard and imported older source before collecting later regression cases. It failed those new cases with that older cached module. This run is retained, not counted as candidate acceptance. The fresh focused run passed all seven stale-setup cases. No case was skipped to hide a failure. - Historical remote gates for `1c31` were terminal and clean: 53 successful contexts and two expected Storybook skips. Main CI [40720536514](https://github.com/paperclipai/paperclip/actions/runs/40720536514) passed. Greptile reviewed exact head `1c31cde0ed199fbf4e41b7649590554f7886dfc2` at 5/5 with no actionable findings; the separate review job also passed, and there were no inline review comments. These results do not qualify `6c18`. - CI source qualification: the chat jobs checked out synthetic merge `ef36746b3d2ec52ba5727dcbe4da265a03421fe4`, which merges the PR into `7d75ae0569e41786640cf09647ae8849411d2f22`. All 20 locally failing case names have explicit passing CI log lines. These passes do not establish standalone local acceptance of `1c31`. The merge adds two upstream Teams tests, so it collects 1,595 chat cases instead of 1,593. Neither backlog/FIFO assertions nor their timeouts changed. - Independent source review found an earlier credential-identity race. The guard and seven deterministic cases resolve that finding. A later integration review missed the separate overlapping-start race. Both RED failures and that review miss are retained. The narrow follow-up repair passed 498 service cases, independent source review and fresh remote CI/5-of-5 review; full local qualification remains incomplete. - Source qualification is ready for merge. Live repaired-draft recovery, delivery and Google approval are not established by these results. The local whole command remains explicitly incomplete. ## Risks - Historical drafts without a matching intent now need an explicit service choice. This is deliberate and prevents silent broadening. - A concurrent credential or method edit now returns 409. Reloading lets the person review the changed setup. - Overlapping starts keep separate requested scope sets. Updating draft intent does not relabel an earlier OAuth state's consent request. - Saved intent is not consent. Provider-reported scopes and current action policies still control discovery and dispatch. - No migration is needed. No existing token or grant is rewritten by saving draft intent. - Default ungrouped launch paths now execute six chat children in sequence and 28 whole-command invocations. Exact coverage, local launcher contracts and all six real default chat partitions passed. A complete uninterrupted local whole-command run remains unqualified because of host capacity. - The retention test-clock repair depends on owned-I/O and zero-duration connection-timer progress. The controlled regression, sensitivity mutant and current exact-tree CI protect the original assertion. It did not reach a completed row in the interrupted local whole run. ## Model Used OpenAI Codex, described by the runtime as GPT-5 based. The exact runtime model ID and context-window size are not exposed to this task. Used reasoning, code execution, browser verification, and independent delegated review. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used, including the runtime metadata limit - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have described the issue in-PR following the bug report template - [x] I have not referenced internal or instance-local Paperclip issues or links - [x] My branch name describes the change and contains no internal ticket or instance identifier - [x] I have run tests locally and they pass — exact clean 82 passed 41 launcher/workflow cases and all 1,816 real default chat cases, full recursive typecheck, normal build and token gates. Full exact-tree PR CI passed. The separate local whole `pnpm test:run` was interrupted by disk protection and remains incomplete; it is not claimed as a local whole pass. - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented risks above - [x] All Paperclip CI gates are green — current exact-head checks passed; the verified CI checkout tree equals clean 82. The separate interrupted local whole command is disclosed above. - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups — fresh exact-82 summary and check passed; the earlier finding is resolved. This is source review, not acceptance. - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
7114582c63 |
fix(ci): track repository inputs across test projects (#18398)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - PR CI selects affected tests before allocating parallel runners. > - Recent audits retained about 96% of test files despite narrow changes. > - Unchanged consumer packages kept full coverage, and source scans confused data fields and message text with runtime loaders. > - File reads need explicit edges that imports alone cannot provide. > - This change follows consumer graphs and known file inputs while retaining unknown effects. > - The AWS CODEOWNERS gate stays in front of runner allocation. ## Linked Issues or Issue Description Refs #18376 and #18362. Narrow source changes still selected most tests because consumer packages kept full coverage and known file reads were treated as arbitrary input. Expected behavior is to retain affected import/file consumers and unknown inputs, then partition only the selected files. The shared storage failure found during the audit was fixed upstream in #18357. The candidate merge ran all 18 managed-skill storage tests successfully in [Linux qualification](https://github.com/paperclipai/paperclip/actions/runs/41140209697/job/124352365547). This branch includes that upstream commit; no additional storage implementation or permanent diagnostic job is included. ## What Changed - Assess every Vitest project in the coarse plan, including unchanged consumer workspaces. - Add dependency edges for statically resolved module-relative filesystem reads and directory listings. Keep unknown paths, writes, escaped APIs, mutable path objects and untracked inputs conservative. - Distinguish data keys, erased types and message strings from runtime globals, loaders and mock bypasses. Retain destructured globals, aliased object reads and mocked filesystem/path resolution. - Normalize `fs` and `node:fs` before recording coverage so a known read cannot hide an unknown one. - Review the Drizzle SQL/schema entry points by hashes of their complete import closures. CommonJS/dynamic loaders reject the fingerprint; drivers and migration readers remain opaque, and dependency updates revoke the review. - Record file input edges and per-test retention reasons in selection audit artifacts. - Add cross-workspace, mixed-import, symlink retargeting, setup-input and dependency-review regression cases. Object-valued URL constants are explicitly opaque with and without mutation. Update the developer guide. - Preserve the AWS CODEOWNERS authorization gate, runner-group restrictions and existing lane ownership. ## Verification - Final head `eedad362e84d6b68f0f048c92db6032e45622c46` passed [complete Actions CI](https://github.com/paperclipai/paperclip/actions/runs/41144233520) on attempt 1. Overall status: 57 successful checks, three intended skips, and one neutral Paperclip review. Both Greptile and Paperclip rated this head 5/5; zero unresolved review threads. - All 17 local selector tests passed: `node --test scripts/ci/test-selection*.test.mjs`. Final [Linux qualification](https://github.com/paperclipai/paperclip/actions/runs/41144233203/job/124365428832) passed all 18 selector/impact tests without skips, plus real workspace selection and cache/routing checks. - Same-tree local comparison against the previous selector at `d331d8481a`: Inbox changes selected 466 rather than 700 files (33% fewer); auth middleware changes selected 2,141 rather than 2,788 files (23% fewer). The sampled plans retained the same 2 and 24 jobs under existing weighted budgets. Final Linux qualification selected 466 files for Inbox and 1,949 for auth/config-test changes; platform-dependent graphs differ, so these are not a Linux before/after comparison. Chat/native Runner lanes retain their existing ownership and are excluded from these counts. - Local `pnpm -r typecheck`, `pnpm build` and `git diff --check` passed. The complete local `pnpm test:run` was attempted and stopped after repeated embedded PostgreSQL startup failures; initdb reported exhausted macOS semaphore capacity. Full Linux CI passed on the final combined tree. - The storage regression was corrected upstream in #18357. [Focused Linux verification](https://github.com/paperclipai/paperclip/actions/runs/41140209697/job/124352365547) ran all 18 managed-skill tests successfully, with no skips; the final full CI run also passed its server lane. ## Risks - Unknown effects must retain their consumers. Regression cases cover mixed builtin spellings, dynamic paths, escaped/mocked APIs, global aliases, reviewed dependency changes and cross-package file reads. - Graph analysis now visits unchanged consumers. Its added work must be included in rollout latency measurements; fewer test files do not by themselves prove shorter end-to-end PR time. - The weighted shard budget deliberately remains unchanged. These samples reduce file execution without reducing allocated jobs. ## Model Used OpenAI Codex, based on GPT-6, with reasoning, repository inspection, code execution and GitHub tools. The exact serving variant and context window are not exposed in this session. No subagents were used. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass — targeted selector suite; full local suite limitation documented above - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Devin FoleyandPaperclip
|
105a67b52c |
Reuse signed master images for npm canaries (#18404)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Release workflows publish npm packages and multi-platform images. > - A master push can finish its image before the npm canary dispatch starts. > - The dispatch currently builds that same source again on two runners. > - This PR verifies and reuses the signed master image by digest. > - Missing images retain the native build path. Channel promotion still follows npm. ## Linked Issues or Issue Description **What existing behavior does this improve?** The Docker workflow dispatched after npm canary publication. **Current behavior** Every canary dispatch builds both architectures, even when master has already published the same source. This consumes runner capacity and delays the image channel update. **Proposed behavior** Resolve the full-SHA master image. Verify its platform index, digest and signed source identity. Retag that immutable digest. Build normally when the image does not exist. **Reason and benefit** Remove duplicate image builds while keeping missing-image recovery and the current npm channel check. A recent duplicate used about 25 runner-minutes. The saving depends on whether the master image exists. ## What Changed - Add a canary-only reuse job before the native matrix. - Require signed provenance from the canonical Docker workflow on master at the exact source and signer commit. Read the proof from GHCR. - Reuse the validated digest or build both platforms when the full-SHA image is absent. Stop on registry and verification errors. - Keep the serialized npm dist-tag promotion after either path. Keep other build lanes independent. - Add failure tests, register them in CI and update release documentation. ## Verification - Passed 603 tests: `node --test '.github/scripts/tests/*.test.mjs' scripts/standard-image-contract.test.mjs scripts/reuse-canary-image.test.mjs scripts/__tests__/release-verify-workflow.test.mjs scripts/cloud-source-verification.test.mjs`. - Passed `actionlint -shellcheck='' .github/workflows/docker.yml .github/workflows/pr-trusted.yml` and `git diff --check`. - Full actionlint reports the same six existing shellcheck findings as the unchanged master workflow. No new findings. - A published master image passed the exact provenance command. This was a read-only check. No images were retagged. - Gitleaks and manual diff inspection found no secrets or private references. - Attempted `pnpm -r typecheck`, `pnpm test:run` and `pnpm build`. They cannot start because this isolated checkout has no installed dependencies (`tsc` and `tsx` are absent). Repository CI passed on `3c46f0152d4db94a7f393bc79f466aa6c2c745e2` after the documented browser-test rerun. All checks are complete. The original failure and prior baseline occurrence remain recorded in the PR comments; its root cause remains unknown. ## Risks - The image can appear just before its attestation. Verification then fails closed. Rerun after the master producer completes. - Reuse preserves the first build's resolved dependencies and tool versions. It intentionally does not rebuild the same source to refresh floating inputs. - A skipped or unfinished master image still needs a full build. No speedup is claimed for that path. - End-to-end reuse and runner savings require observation after merge. No live publication was performed. ## Model Used OpenAI `gpt-6.1-sol`, through the Codex harness in T3 Code, with xhigh reasoning effort and shell/code tools. The context window size is not exposed by this runtime. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
56957f2bd9 |
fix(ci): isolate test selection and require current CI CODEOWNERS (#18376)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - PR CI selects affected suites and test files to reduce feedback time. > - A selected test inherited the parent selection when it launched an independent runner fixture. > - Ordinary source filenames containing config or setup also forced broad coverage. > - AWS runner access must depend on current CI CODEOWNERS, not only a stored numeric roster. > - This pull request fixes both selection problems and verifies owners before AWS allocation. > - Contributors retain GitHub-hosted CI. Authorized maintainers retain AWS test and final verification jobs. ## Linked Issues or Issue Description Refs #18362 and #18348. The selected server run for #18369 failed in `vitest-chat-shards.test.ts` because its temporary fixture inherited `PAPERCLIP_CI_TEST_SELECTION` and attempted `git rev-parse HEAD` outside a Git repository. #18366 could not use graph selection because `session-config-metadata.test.ts` matched the broad filename rule. The AWS setting was already enabled. The reviewed outsider PRs correctly used hosted runners; a maintainer PR used AWS, including final verification. This change strengthens that boundary rather than enabling AWS for outsiders. ## What Changed - Remove selection settings from Vitest child environments after the wrapper validates the selected shard. - Match conventional Vite/Vitest configuration filenames rather than any occurrence of config or setup. Keep graph tracking for actual setup and configuration dependencies. - Bind rerun actor lookup to the exact run attempt. Require the author, event sender and rerun actor to be current CI CODEOWNERS and members of the existing numeric allowlist. Read the resolver and CODEOWNERS from one protected master revision. - Run metadata authorization on GitHub-hosted capacity before any AWS allocation. Keep authorized test and final verification placement on the AWS fleet. - Add a real selected-server regression that launches nested chat fixtures, negative authorization cases, and qualification for a real config-named test. - Update the developer and profiling guides. - Retire eight legacy PR workflow SHA authorizations and one deleted qualification-branch reference from the live `paperclip-public-pr` runner group. The remaining three references are the trusted PR workflow and the existing full-stack/protocol evaluation workflows on master. No active PR run referenced the retired versions. ## Verification - Selection regression tests: 13 passed, including the real nested chat runner. - Routing, planner and profiling tests: 100 passed, including 69 routing cases. - `actionlint` passed for both changed workflows. `git diff --check` passed. - Confirmed protected master still requires CODEOWNER review. A live lookup found four CI owners, with three of six numeric roster entries eligible and three denied by the new owner check. - `pnpm -r typecheck` and `pnpm build` passed. The full local Vitest suite was stopped after repeated embedded PostgreSQL initialization failures; it did not pass. On final revision `0b40ae15b7996603e405edd4157a5c337e0405c7`, both Linux qualification jobs and all required CI gates passed. The main PR workflow completed with 48 successful jobs and one intentional skip, including final verification. [CI run](https://github.com/paperclipai/paperclip/actions/runs/41126881716). Greptile reviewed the final revision at 5/5 with no inline findings. - Shard and workflow contracts: 44 passed. CI first caught an outdated assertion that required the routing job to use AWS; it now requires hosted authorization, and the policy job passes. - A read-only live gate exercise preserved the existing merge-freshness denial when master advanced beyond the available merge ref. It allocated no runner and was not a successful admission proof. ## Risks - The short metadata authorization job now depends on GitHub-hosted queue capacity. This prevents a stale roster from allocating AWS before current CODEOWNERS are checked. Authorized test and final verification jobs still use AWS. - Identity/API failures fall back to hosted CI. No caller can override the trusted owner list with PR contents. - Configuration dependencies still require conservative graph handling. Existing setup dependency, opaque I/O, stale manifest and wrong-lane tests remain in place. - The runner group is restricted to selected repositories and selected workflows. Its workflow allowlist was verified and narrowed through the GitHub settings UI. Legacy pinned callers must update to the current trusted workflow before using AWS; active PR runs used master references. The current PR workflow authorization and evaluation workflow authorizations remain in place. ## Model Used - OpenAI Codex, based on GPT-6. The exact serving variant and context limit are not exposed in this session. Used reasoning, repository inspection, code execution and GitHub tools. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass — targeted suites passed; the full local suite hit the PostgreSQL startup limitation documented above, and full Linux CI passed - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
76fde891cc |
Add bounded recovery file inventory before server cutover (#18347)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Accounting, credential recovery, and sandbox cleanup can retain legacy files on an instance volume. > - Replacing that volume while files remain can lose unresolved work. > - There is no verified production backlog, so an empty installation needs no migration. > - The recovery owners must decide whether each record is valid and settled. > - This pull request adds a bounded file inventory and an optional cutover gate. > - The benefit is a safe check that preserves obligations without reading secret material. ## Linked Issues or Issue Description **Problem or motivation** Operators need to check the actual affected instance directories before replacing a web server. A skipped corrupt file or failed read must not appear clear. **Proposed solution** Add a Node script that inspects only the three legacy recovery directories for an explicitly chosen instance. Default mode reports counts. The optional check mode blocks when files or uncertainty remain. **Alternatives considered** A bulk migration system has no established backlog to migrate. Reimplementing the owner parsers could expose credentials and still would not prove settlement. This check never reads file bytes. **Roadmap alignment** This is a small operational prerequisite for safe cloud server replacement. It does not add a recovery backend or change resource ownership. Related recovery-owner changes: #18354 (accounting), #18352 (credentials), and #18357 (sandbox ownership). This inventory remains independently usable with existing files. ## What Changed - Add a pure classification policy and a bounded filesystem boundary. - Add `node scripts/recovery-cutover.mjs --instance-root /absolute/instance` with optional `--check`, directory override, and entry budget. - Report observed lower-bound counts and safe status summaries. Do not print paths, filenames, identifiers, raw errors, or file contents. - Add one dependency-free CLI acceptance suite and a Vitest bridge so the same cases run in normal server CI. - Document existing owner entry points, override selection, stopped producers, and retained-volume requirements. ## Verification - `node --test scripts/recovery-cutover.test.mjs`: 9 cases pass. The suite starts the actual CLI with synthetic legacy directories. - Pending receipts, nested credential files, orphan records, malformed bytes, temporary files, unreadable files, unsafe paths, symlinks, read failures, bounds, and configured overrides all block safely. - Empty or absent recovery directories clear. Source file bytes remain unchanged. Report-only mode and check-mode exit codes are tested. - `pnpm exec vitest run server/src/__tests__/recovery-cutover.test.ts --pool=forks --maxWorkers=1`: the normal CI bridge passes and runs the same 9 cases. - Cached frozen dependency installation, script syntax checks, and `git diff --check` pass. The lockfile is unchanged. - Independent source review found the read-only boundary sound. - [Complete current-head CI](https://github.com/paperclipai/paperclip/actions/runs/41097614666) passed at `bf5cc7f067041c10cd0294ff26756678da943976`: 53 successful checks and two intended skips, including typecheck, full tests, build, and E2E. Greptile is 5/5 with no review threads. - One failed-job retry was needed for an existing heartbeat timing failure in server shard 4. The relevant heartbeat source and tests match the base branch. Local diagnostic runs had 61 pass/2 fail and 7 pass/1 fail, with changing timing failures; neither run is claimed as green. The bounded retry and all failed evidence are [recorded here](https://github.com/paperclipai/paperclip/pull/18347#issuecomment-6529629804). No heartbeat code or test was changed. ## Risks - The scan is a point-in-time inventory. Stop producers for the final gate. It is not an atomic drain certificate or protection against hostile path mutation. - All files remain unverified. The check cannot certify validity or Postgres settlement. Existing owners retain that responsibility. - The scan does not inspect other instances, volumes, Runner delivery, or database obligations. It does not establish that production cutover is ready. - No migration, deletion, decryption, or replay command is added. Symlink paths are rejected, including system aliases; use canonical absolute paths. ## Model Used - OpenAI `gpt-6.1-sol`, with reasoning, tool use, and code execution. Context window size was not supplied. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
9f524ec1a7 |
fix: persist native Runner usage before transport acknowledgement (#18354)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Agent execution must preserve usage so Paperclip can price work and enforce budgets. > - Remote Runner execution can outlive its web receiver. > - The receiver previously acknowledged Runner events before its later usage recorder saved the accounting receipt. > - This pull request saves required native usage receipts before that transport acknowledgement. > - A replacement receiver can accept the same retained report and create one ledger charge. ## Linked Issues or Issue Description **What happened?** Native remote Runner usage could leave its durable outbox before the web receiver saved its accounting receipt. The later recorder required a local receipt directory. A replacement web receiver could lose the only remaining report when the database write failed. **Expected behavior** The Runner retains unaccepted structured usage. Postgres owns accepted receipts. Repeated delivery preserves one receipt for each source report and one charge for the run. Missing usage keeps accounting pending. **Steps to reproduce** 1. Start a supported built-in native remote Runner. 2. Make the receipt database insert fail after the provider reports usage. 3. Replace the web receiver and its local directory. 4. Reconnect the same Runner and inspect the saved receipt and ledger. **Paperclip version or commit** The initial investigation used `ebd05e9f4c88c66af7473db86bbaacfd1544aa0a`. **Deployment mode** A web server with a remote native Runner and shared Postgres. Related work: #18327 owns output logs and provider traces. This change has no dependency on its S3 export path. #18031 covers interrupted Codex receipt recovery. #18005 covers legacy spool reads. #17516 covers Codex stop-boundary retention. ## What Changed - Add native producer receipt delivery through the existing Runner callback before transport acknowledgement. - Reuse `run_usage_receipts`, stable source sequences, and deterministic report identities. Reject conflicting replay and retry a lost commit response with the same receipt. - Restore per-turn usage from structured receipt facts. Preserve provider turns that share one PRP envelope turn. - Bind retained warm transport callbacks to the current run authority epoch, allowing old attachment control acknowledgements while preserving accounting ownership. - Persist Codex session baselines and cumulative totals in the existing provider checkpoint. Mark qualified whole-run deltas, preserve them across resume/restart/attachment, and leave old unmarked reports incomplete and unpriced. - Normalize the bounded final usage tail before the terminal event. Retain later increases for explicit reconciliation, block further attachment, and fail an affected successor without changing its baseline. - Complete runs from the accepted immutable receipt and its exact saved rate-card amount; preserve existing live-budget enforcement. - Skip the local receipt recorder for supported built-in native remote Runner execution. Keep local, legacy, and external Dot/Muse behavior unchanged. - Keep pricing, reservations, ledger settlement, source fencing, and budget holds with their existing owners. - Put the actual Runner integration test in the existing Rust-cached native test lane. Build before opening Postgres and give cold compilation its own bounded budget. ## Verification - Fresh Greptile review is 5/5 on `e7376a50c8162770da1d581ecf25c8cffac64520`, with no unresolved findings. Parent and independent peer review approved the final changes. - Real PostgreSQL 18.6 and a real Rust Runner exercise failed receipt insertion, receiver-directory loss, replay from the same producer, and exactly one 25-cent ledger charge. The actual accounting spool path stays absent. Optional logs/traces are disabled. The provider is controlled; no paid request is made and the fixture never pre-saves the expected receipt. - Producer checks passed 32 usage/provider tests,37 terminal controls, corrected completion, all 3 previously failing exit cases, and both full transport goal-continuation cases. Rejected active-goal completion retains its 120-token tail before failure. Baseline 100→successor150→33 duplicate old snapshots→old120 retains explicit incomplete reconciliation evidence and positive budget exposure. - Heartbeat executor checks passed 982/982. PostgreSQL tests preserve exact 1.002 USD pricing, the 100.2-cent ledger amount, budget holds, source takeover, partial/final receipts, cost-only corrections, equal rate-card replay, and stale-completion interleaving. Newer incomplete usage remains pending rather than settling an older snapshot. - Runtime replay and strict usage-schema checks passed 2/2; generated contracts and Rust formatting pass. Test-partition contracts passed 32/32. - [Final combined-head evidence](https://github.com/paperclipai/paperclip/pull/18354#issuecomment-6531526158) records full workspace typecheck/build and 41/41 accounting tests against all five cloud PR heads, including this final accounting commit. It includes commands and evidence limits. - Exact-head CI is green: 53successful checks, 6intentional skips, and the legacy Snyk status passed. Typecheck, build, Runner, server, browser, and aggregate gates pass. Runner Vitest1/2 initially failed an unchanged mocked Antigravity login fixture; its focused case passed locally, then one unchanged job retry passed 131 files/1995 tests, including all 11 Antigravity tests. The initial failed attempt remains recorded in CI. Initial local attempts exposed the host SysV semaphore limit and corrected harness path, IPC readiness, and queue-counter defects. A temporary test loader selected an existing POSIX-semaphore PostgreSQL 18.6 build without changing repository dependencies. An optional expanded debug provider run was interrupted at its unrelated 8192-turn rollover stress case; no full-file local pass is claimed. Required failure cases were run directly, and the full Runner CI remains required. ## Risks - The guarantee covers built-in native remote Runner execution. Local native execution, legacy adapters, external Dot/Muse execution, and server-originated model calls retain their existing contracts. - This does not provide traffic switching or general web-server failover. The existing controller identity, checkpoint, and generation gates still apply. - The receipt JSON gains optional structured producer facts. No schema migration or accounting S3 spool is added. Text output and raw trace payloads are excluded. - Codex accounting authority requires a Runner artifact built from this provider baseline implementation and qualified by digest. Update receivers and strict usage-schema readers before releasing that Runner; PRP does not negotiate this payload field. Existing unmarked Codex reports remain pending/unpriced. No deployed Runner artifact is claimed fixed. - Usage arriving beyond the bounded terminal tail may require manual reconciliation or a fresh session. Its evidence remains in the existing producer checkpoint/outbox; no adjustment ledger or recurring scan is added. ## Model Used OpenAI `gpt-6.1-sol` assisted with the implementation, tests, and documentation through code execution and repository tools. The context window size is unavailable. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
9e1735ec17 |
test(evals): add a fixed TestOps AWS chat feature route (#18322)
## Thinking Path > - Paperclip lets people manage AI agents and inspect their saved work. > - Chat feature evaluations test a real request against saved tasks, audit records, links and reload behavior. > - The current suite has Codex and OpenCode cells but no dedicated runnable DeepSeek 1017 backlog plan in TestOps. > - TestOps must select the real case before the workflow can access a model key. > - This PR adds one fixed, default-branch AWS route for native OpenCode and DeepSeek 1017. > - The route keeps original evidence private and sends only a verified result and approved screenshots to TestOps. ## Linked Issues or Issue Description **Subsystem affected** Product E2E evaluation definitions and GitHub Actions delivery. **Problem or motivation** A dedicated TestOps plan cannot run the existing backlog chat journey with native OpenCode. A result must prove saved work and cleanup rather than report that a model call completed. **Proposed solution** Add one OpenCode backlog cell and one fixed public workflow. TestOps selection validation runs in a token-isolated job before builds or inference. The workflow retains private originals and reports the bound outcome to the selected TestOps case. **Alternatives considered** A private-repository wrapper would need a new cross-repository GitHub write credential. The new route uses the existing TestOps GitHub integration instead. Existing manual candidate approval paths remain separate. **Roadmap alignment** This extends the existing feature evaluation infrastructure. It changes no production feature, global runtime default or model default. Current main #18319 supplies the v2 grader, delegation queue accounting and final screenshot capture; these improvements are preserved. Related product work: #16814 and #17126 concern backlog handling. This PR changes neither product implementation and does not claim those changes are required. ## What Changed - Add exactly one native OpenCode backlog cell with DeepSeek 1017 under the separate `runner-opencode-deepseek-1017` profile. Preserve all four existing Codex and production OpenCode cells and their models. - Add a default-branch-only workflow with fixed evaluation, zero-result registration and one plan-launch modes. Require GitHub actor checks and authenticated singleton selection before inference. - Add a separate shipped-main evidence authority. Reuse the existing strict private storage, scanners, inventories and scoped AWS sessions. - Bind sanitized results, generated JUnit and producer-approved screenshots to the exact source, run, attempt and case. Keep failed grades and cleanup failures failed. - Add TestOps API, CLI and result adapters with pure controls and normal unit-test wiring. Keep the reporting token out of the evaluated process. - Verify the current server case result and attachment metadata after upload, without masking an earlier failed upload or evaluation. - Apply only the existing reviewed lock correction, verify unchanged tracked dependency inputs and bind the lock receipt across the compiled-output handoff before frozen installation. - Bind the initial administrator launch to one approved workflow run number and live GitHub source/actor metadata before TestOps authentication. Use bounded REST case reconciliation and only the reviewed existing GitHub integration. - Make two existing CI tests observe their intended state deterministically: a fake clock for copy-feedback lifetime and a completed-pass barrier for retry-cancellation batches. Production behavior and the original assertions remain unchanged. The two test-only fixes are intentionally included because their observation boundaries blocked this delivery’s required normal CI. They preserve every original assertion and change no production behavior. The original failed runs remain recorded, and the exact causes of those observations remain unknown. ## Verification - 157 focused profile, catalog, oracle, filesystem admission, export, private-path and workflow controls passed on the current-main composition. The existing stock OpenCode models and the new exact DeepSeek 1017 selection are checked separately. - 44 API, CLI, filesystem and post-upload readback controls passed, including foreign selection, changed hashes, unsafe files, failed outcomes, expired readback and no-resend behavior. - The new exact private session and trace layout controls produced four genuine parent failures and pass with the selected-path extension. The frozen 244-case candidate approval registry remains unchanged. - The lock guard has seven filesystem controls plus seven existing canonical-owner controls. The unrestricted old workflow failed its actual policy calibration; the corrected frozen-install route passed. - Eval-owning TypeScript checks, workflow lint and Git whitespace checks passed. The separate UI test correction was checked with the focused jsdom unit suite. The server test barrier has pure wrapper and real scheduler controls; its DB-backed route was not run locally. No additional local typecheck/build was run for these test-only changes. - Full native builds, provider execution, app/database behavior and TestOps registration are reserved for GitHub Actions/AWS. They have not run locally. - The initial normal CI run failed the new-workflow policy allowlist and a heartbeat scheduling assertion. The policy fix has an actual parent red/current green control and rejects five unsafe route variants. Heartbeat source is unchanged and its failure cause remains unknown. The original run is retained. - The launch-fence actual caller fails against the published parent and passes the five own-run contradiction controls before authentication. - The prior `a5aa359d` normal CI completed successfully. The complete `f9e1f289` run passed all leaf jobs except an existing copy-feedback UI assertion, plus its dependent verification gate. The four owning UI files were unchanged from the earlier passing head and actual tested merge; the exact scheduling cause remains unknown. - The correction changes only that test to actual React `act` and a scoped fake clock. All eight component tests pass; early-hide and never-clear source variants each fail the intended lifetime assertion. Production clipboard/card behavior is unchanged. - The `2f27f3fc` run passed the corrected UI shard and every leaf job except the existing retry-cancellation intermediate-count assertion, plus its dependent verification gate. The seven relevant owners match the earlier passing run and historical source comparison; the exact CI scheduling cause remains unknown. - The test-only barrier observes a persisted completed worker pass before asserting the unchanged 20-item cap and two-company fairness, then releases it for the original 29-item drain and 180-second idle check. Two extracted-wrapper controls and three actual scheduler controls pass. The real DB-backed test awaits AWS qualification. - On `836e80ac`, both repaired test shards passed in AWS. The original attempt retained a separate E2E cleanup HTTP 500 from a PostgreSQL lock timeout and one runner shutdown. No production repair or precise contention/shutdown cause is claimed. - Exactly one targeted retry of E2E shard 3 and its aggregate gate passed in attempt 2; exactly one recovery of the interrupted server shard and its verification gate passed in attempt 3. Other passed jobs were carried forward. The final normal run is successful. No whole-workflow retry or evaluation/model retry was launched. - Both literal required checks (`ci / verify` and `ci / e2e`) are successful on the exact final source `836e80ac`. Greptile is 5/5 and no review threads remain open. Actual TestOps registration and the OpenCode/DeepSeek feature result remain unqualified until their separate executions. ## Risks - Actual zero-result registration, tenant API shapes and public-repository dispatch access must be checked before the first evaluation. Missing or ambiguous metadata stops the route. - The real OpenCode/DeepSeek backlog result and Allure outcome remain unqualified until that run. No passing result is seeded. - One human query, one accounted work run, 120-second turn, 420-second attempt, 300-second settled-state bound and zero retries remain unchanged. Setup usage is outside work accounting and stays unknown. - The reporting token was configured through the separately reviewed protected credential bootstrap. It is unavailable to AWS evaluation jobs. - Metadata reconciliation is capped at ten pages, 1,000 cases and four parallel detail reads. Missing or inconsistent metadata stops before inference. The initial administrator launch counter must be created once after registration; it does not restrict normal TestOps Run actions. - Existing manual workflows, historical evaluation grades and candidate approval records remain unchanged. The original CI failures are preserved; the test corrections do not establish retrospective causes. ## Model Used OpenAI GPT-6.1 Sol (`gpt-6.1-sol`) with reasoning and tool use. The runtime did not expose its context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
3b84c89885 |
fix(ci): use complete merge diffs and allow audited documentation (#18362)
## Thinking Path > - Paperclip manages AI agents and needs reliable checks for each contribution. > - CI selects affected suites and then affected Vitest files before allocating runners. > - Recent runs compared a tested merge against a stale event base. > - That included unrelated master changes and sometimes reached the compare API file cap. > - Ordinary documentation also forced every suite and invalidated test caches. > - This PR reads complete merge trees and exempts a small set of audited prose inputs. > - Tests retain conservative coverage when the diff or an input is uncertain. ## Linked Issues or Issue Description Refs #18348 and #18214. This is a follow-up to the merged affected-test selector. Related CI performance work: #16513 and #16500 address setup and suite execution. This PR changes diff collection and input classification. ## What Changed - Fetch the exact tested merge and its parents into a temporary bare repository. Read the complete tree diff against the actual first parent. Validate the second parent against the event PR head. Never execute PR code in the routing job. - Carry the immutable base, merge and PR-head identities into the plan. Validate the same complete diff again before file selection. - Use one prose policy across the planner, Vitest selector and task cache adapter. Exempt only `doc/DEVELOPING.md`, `doc/DATABASE.md`, and Markdown under `doc/plans/`. - Keep the uncached typecheck and CLI guidance guard for the first two documents. Keep policy checks for historical plans. Keep existing coverage for all other docs, runtime skills, fixtures and unknown inputs. - Extend candidate-workflow qualification to exercise mixed source and prose changes. ## Verification - Passed: 15 planner and Git routing tests, including a 353-file diff, stale base, rename, deletion, wrong parent and fetch-failure cases. - Passed: 83 routing, Vitest graph, exact selection, workflow and real Turbo cache checks. - Passed: `pnpm -r typecheck` and `pnpm build`. - Full local `pnpm test:run` was attempted and stopped after repeated embedded PostgreSQL startup failures in native-session and Muse tests. It did not pass or complete. The exact cause remains unverified. The changed files do not include those tests or their database setup. Linux CI results are recorded below. - Passed: real mixed Inbox/source/prose graph qualification selected 699/911 workspace test files, preserving the source-only selection. - Passed: both GitHub CI qualification jobs, including real mixed source/prose Vitest selection. Greptile reviewed commit `10ac6f5ec9` at 5/5 with no findings. - First full CI attempt: the [chat shard](https://github.com/paperclipai/paperclip/actions/runs/41111003419/job/124253342684) received a runner shutdown signal. The [Runner shard](https://github.com/paperclipai/paperclip/actions/runs/41111003419/job/124253341271) failed the unchanged Antigravity authentication success assertion. All 11 cases in that file passed when run directly. That Runner lane does not use the changed selector/cache path. The exact assertion-failure cause remains unverified. The chat shard passed on the one retry. The authentication assertion also passed on retry, but the same Runner shard failed a different unchanged test: `native-distribution-integrity.test.ts:242`, "retains spawned native bytes until the exact child exits", exceeded 5000ms. The [retry log](https://github.com/paperclipai/paperclip/actions/runs/41111003419/job/124259174201) preserves the failure. The Runner shard and required verify aggregate are still red; all other checks passed. All 33 native-distribution tests then passed in a direct local run. This shows recovery in isolation, not the cause of the Linux timeout. No further blind reruns were issued. - Replayed 15 real recent PR merge commits through a blob-filtered bare fetch. Nine needed a corrected base. The four API comparisons that returned 300 filenames became complete diffs of 3, 57, 66 and 54 files. The 3-file case (#18326) becomes eligible for graph selection. Most other samples still correctly use full CI because they include inputs outside this narrow prose policy. - Cache proof: prose edits preserve a product-test cache hit; changing runtime API documentation invalidates it. ## Risks - Prose exceptions must be removed or revised if product code starts reading those files. The allowlist is deliberately small. Runtime docs and skills remain covered. - A network or Git failure keeps the full-suite fallback. Unexpected merge topology also keeps full CI. Stack and force-full policies remain in place. - This does not claim an end-to-end speedup yet. Many mixed-content PRs still require full CI. The routing fetch reads only shallow commit/tree objects; it does not check out or execute PR files. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, repository tools and code execution. The exact served model ID and context-window size are not exposed to this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
52e846f2f3 |
perf(ci): select affected Vitest files before allocating test runners (#18348)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Its CI must catch regressions without running unrelated work on every pull request. > - The existing task planner selects package lanes, but each selected lane still runs all its test files. > - Vitest can find related imports, but that graph misses tests that read files or launch processes. > - This pull request combines the real Vitest graph with conservative dependency checks before CI allocates test runners. > - The result runs fewer test files and can allocate fewer jobs while retaining tests with unknown dependencies. ## Linked Issues or Issue Description Refs #18214. The repository and PR search found no duplicate file-selection implementation. **What existing behavior does this improve?** The general and serialized Vitest lanes in pull request CI. **Current behavior** An affected package schedules its complete test lane and its fixed shard count. **Proposed behavior** For source-only diffs, select related tests plus conservative filesystem, process, network and opaque consumers. Repartition those files within their existing lane owners. Keep the complete original plan when selection cannot be trusted. **Reason and benefit** Avoid unrelated test execution and allocate fewer runners when the selected workload fits fewer existing job budgets. Report file counts and job counts separately. **Breaking changes** No product or API changes. The set of PR test files and jobs changes for eligible diffs. Full CI and the kill switch retain the original coverage. ## What Changed - Add a separate selection job on the gate-approved test fleet. The metadata gate does not load PR code. - Inspect real Vitest/Vite dependency edges. Retain opaque consumers and unchanged downstream projects. Respect only complete local mock factories. Reject unknown setup dependencies and source mutations during dependency setup. - Declare `oxc-parser` directly for AST inspection. This exact version is already a transitive dependency in the checked-in lockfile; the direct declaration avoids relying on pnpm's internal dependency layout. - Produce a merge-SHA-bound manifest with exact lane ownership and duration-aware partitions. Preserve serialized execution, chat shards and native Runner ownership. - Execute exact Vitest test specifications. Positional substring filters cannot accidentally run a similarly named file from another shard. - Use the original matrix if selection fails or times out. Fail selected jobs if their published artifact is missing or invalid. - Save the uploaded artifact ID in the executable plan. Failed-job retries reuse that artifact even when GitHub increments the attempt number. - Hash selected assignments into Turbo cache inputs. A subset cannot satisfy a full run or another subset. - Add graph, workflow, real-runner and real-Turbo qualification. The regression fixture proves a filesystem contract fails even when raw Vitest selection misses it. - Document selection limits, audit artifacts and `CI_VITEST_SELECTION=off`. ## Verification - All 138 affected routing, sharding, workflow, manifest, real Vitest and real Turbo checks passed under Node 24 on the final implementation. Additional release workflow checks passed. Cases cover opaque setup, remocking, generated inputs, Node's builtin loader, artifact retries, and exact execution of similarly named files. - Repository-wide `pnpm -r typecheck` and `pnpm build` passed before the clean rebase onto current master. The full remote typecheck and build are required on the current head. - `actionlint` passed for both changed workflows. `git diff --check` passed. - The Linux qualification run passed and selected 699 of 911 workspace test files for an Inbox change. UI graph construction took 16.1 seconds. This still uses two jobs. Synthetic partition coverage proves a 15-of-120 selection uses two of twelve jobs without dropping or duplicating files. - Both selected workspace shards passed locally. After tightening execution to exact specifications, the real colliding UI test ran as exactly one file with two passing assertions; the real server auth middleware test ran as exactly one file with 28 passing assertions. - The server probe selected 1,158 of 1,260 general files and 161 of 163 serialized files for an auth middleware change. Those conservative selections retain 12 general and nine serialized jobs. The server's temporary-home setup exception is pinned to reviewed content hashes; changed setup fails closed. - Full `pnpm test:run` has not been run locally. The full PR test matrix is required because this change modifies global CI inputs. Final-head CI is pending at runner allocation, and both qualification jobs are queued. Other recent repository workflows are also queued. Superseded runs of this PR were cancelled; this PR remains in draft until the required checks execute. - Greptile reviewed final commit `31b484fc5b412da245cffb52a8ac8977228c5651` at 5/5 with no unresolved findings. The earlier qualification run passed; it does not replace checks on the final commit. ## Risks - Dependency analysis adds setup and graph time. Fewer files do not guarantee a shorter critical path. CI summaries expose the actual file and job counts. - Static dependency graphs cannot describe all runtime inputs. Unknown inputs retain tests or use the original plan. The pure-package allowlist requires conservative review. - The trusted workflow runs from master. This PR exercises the new selector in the qualification workflow; ordinary PR selection starts after merge. Set `CI_VITEST_SELECTION=off`, use `full-ci`, or select full/shadow mode to disable narrowing. - The lockfile remains managed by the existing repository actions and is not changed here. ## Model Used OpenAI Codex, based on GPT-6, with repository tools and code execution. The exact serving model ID and context window are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandClaude Opus 5.5
|
abc270ee4c |
fix(runner): attest upstream releases and bind ACPX uniformly for every harness (#18325)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Native ACPX runs admit a provider only when its saved profile and `commandDigest` match the current profile exactly. > - For Copilot, Grok and Pi, `commandDigest` hashed dozens of Paperclip source files. Claude, Codex, Cursor and Antigravity hash only their upstream release. > - Any change to a bound source therefore advanced these profiles. #18212 alone moved Copilot from v88 to v91 and Grok from v15 to v19. Each bump reset in-flight sessions and added immutable identity fixtures (about 160 on master). > - #18237 added the provider pack `contractDigest`, which now guards Runner source compatibility at the pack boundary. > - This pull request changes the Copilot, Grok and Pi digests to upstream release attestations, as the other providers already use. > - It also binds the shared ACPX layer the same way for every ACPX harness: the profile's `acpxVersion` and the pack `contractDigest`, which now covers the Runner's ACPX patch. > - The benefit is that Runner source changes no longer reset these sessions or need new profile versions and fixtures. ## Linked Issues or Issue Description Refs #18237, #18212, #17972. **What happened?** A Runner change that touched a bound source file changed the Copilot, Grok or Pi `commandDigest`. The author had to advance the profile version, regenerate an identity fixture and keep every old fixture. On deploy, saved sessions for that provider failed the exact admission check and started fresh. **Expected behavior** A provider's `commandDigest` changes only when its upstream release changes: an upstream version, a platform closure, the ACPX patch or the Grok launcher. Runner source compatibility is the job of the pack `contractDigest`. **Steps to reproduce** 1. On master, change a comment in `src/drivers/acpx/runtime-host.ts`. 2. Run `node packages/paperclip-runner/scripts/grok-profile-identity.mjs --check`. 3. The check fails. The Grok profile must advance, and the next deploy resets Grok sessions. ## What Changed - Add `packages/paperclip-runner/scripts/acpx-release-attestations.mjs`. It computes the release attestation for each provider: - Copilot and Pi: a canonical-JSON hash of the profile version, upstream package versions and the platform closure pins (Pi also binds its Node version). - Grok: the SHA-256 of `launcher.cjs`. Admission already hashes these bytes. - The attestation binds no Paperclip source, shared ACPX layer, absolute path, pack manifest digest or install layout. - Bind the shared ACPX layer once for every ACPX harness (Claude, Codex, Copilot, Cursor, Grok, Pi, Antigravity): - Add the Runner's ACPX patch (`patches/acpx@<runner acpx version>.patch`) to the pack `contractDigest`. A patch change now retires older image packs for every harness. Other packages' ACPX patches are not read. A build context without the Runner's patch fails. - Remove the ACPX patch hash from Cursor's release declaration (`cursor-contract.json`) and advance Cursor once to profile 17. Copilot, Pi and Cursor no longer bind ACPX in their own digests, like Claude, Codex, Grok and Antigravity. - Saved inputs for every harness already must match the current `acpxVersion`. - Claude and Codex digests do not change. - `generate-acpx-profiles.mjs` writes these digests to `acpx-profiles.json`. `check:acpx-profiles` refuses a stale digest. - Move the Copilot closure pins to `copilot-closure-pins.ts`, so that the generator can import them. `copilot-installation.ts` re-exports them. - Advance each profile once and then freeze it: Copilot 92, Grok 20, Pi 37. A version change at the switch keeps the rollout on the same path as every earlier profile bump. - Replace the historical version lists in `profile-compatibility.ts` with a range. Every revision from 1 through the current one decodes. Claude and Codex still decode revisions 1 through 5. Launch and recovery admission still require the exact current profile and digest. - Delete: - `scripts/copilot-profile-identity.mjs`, `scripts/grok-profile-identity.mjs` and its test - the Grok launcher source stamp - 159 Copilot, Grok and Pi identity fixtures, including three unreferenced `copilot-mainline-v*` fixtures. The Cursor fixtures stay. - Rewrite the affected tests as behaviour checks: - Each provider admits only the exact current profile. Every earlier version decodes and is refused at admission. - The last source-bound digests, and a changed digest at the current version, are refused by recovery and by the backend factory. - The Rust Pi descriptor test uses the generated profile instead of fixtures. - New `acpx-release-attestations.test.mjs`: a relocated copy of the inputs gives the same digests, each upstream input changes only its own providers, and the generator refreshes and checks the manifest. - Remove the fixture exceptions from `.dockerignore`. Add the attestation script and the generator to the Daytona image content inputs. - Update the Pi install probe and the Rust integration constant for the new Pi digest. - Update the docs to describe the current attestation model: `doc/grok-native-runner.md`, `doc/RUNNER-INTEGRATION-PLAYBOOK.md`, `doc/observability.md`, `packages/paperclip-runner/docs/durable-recovery.md`, and the Copilot, Pi and rich ACP architecture reports. Each report now opens with the current profile identity; earlier source-bound profiles remain as dated chronology. No doc references a deleted file. ## Verification - In `packages/paperclip-runner`: - `node --test scripts/acpx-release-attestations.test.mjs scripts/contracts/contract-fingerprints.test.mjs`: 19 tests pass, including a Runner ACPX patch edit that changes the contract digest, other patch edits (including `acpx@0.12.0` for `adapter-utils`) that do not, and a build context without the patch that fails. - `node scripts/generate-acpx-profiles.mjs --check` passes. - `vitest run src/drivers/acpx src/backends src/contracts`: 1647 tests pass, 9 skipped. - `tsc -p tsconfig.json --noEmit` passes. - From the repository root, `vitest run --config tests/runner-e2e/vitest.config.ts tests/runner-e2e/daytona-image.test.ts`: 11 tests pass. - In `server`, the 8 test files that use ACPX profiles, pack digests or `contractDigest`: 1617 tests pass. - Not run locally: the Rust tests and `cargo fmt`, because this machine has no Rust toolchain. CI runs them. - Before this change, a synthetic upgrade check built two Runner revisions that differ only in shared source. A Claude checkpoint saved under the first revision rebinds under the second, and a changed Claude `commandDigest` starts fresh. This confirms that the provider JSON has no other build-dependent field. ## Risks - **One reset at rollout.** The Copilot, Grok, Pi and Cursor digests and versions change once. Saved sessions for these providers start fresh on their next run, as after any earlier profile bump. After this change, Runner source changes no longer cause this reset. - **Lost "source changed, so requalify" signal.** The old digest forced a new profile version when bound source changed. Nothing enforced requalification: Copilot was last qualified at v47. Pack-boundary contract changes are now caught by the pack `contractDigest`. - **Image packs are retired once for every provider.** Adding the ACPX patch changes the pack `contractDigest`, so the controller uploads its bundled pack (about 16 s per remote start) until images are refreshed. Claude and Codex sessions are not reset: their digests and versions do not change. - **Grok launcher and closure pins.** The Grok digest covers only the launcher bytes. Grok's runtime version is still checked separately at admission, as before. - **Grok digest equals early Grok profiles.** The launcher bytes have not changed since profile 1, so the new Grok digest equals the published profile 1 digest. The profile version still separates them: launch, recovery and the backend factory require profile 20. An image pack from that period has no `contractDigest`, so the pack gate refuses it. `remote-provider-pack-verification.test.ts` now covers this case. - **Docker context checks** run `generate-acpx-profiles.mjs --check`, which now imports `.ts` pin files through Node type stripping. `materialize-pi-distribution.mjs` already relies on this on Node 24. ## Model Used - Claude Opus 5.5 (`claude-opus-5-5`) in Claude Code, with tool use (shell, file edits, local test runs). ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
Nicky Leach
|
fba0ab8162 | fix(runner): preserve GitHub helpers and Codex probe policy (#18212) | ||
|
DottaandPaperclip
|
99ec767a42 |
Fix forced CI execution and align OpenCode shell fixture (#18230)
Use one explicit Turbo cache mode for forced CI execution and preserve cache trust boundaries. Match OpenCode’s ignored stdin in the relocated Linux shell-hook fixture while retaining strict OOM and parent-isolation assertions. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
f5dcc505c0 |
perf(ci): select affected suites and cache task results (#18214)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Pull request CI checks the server, UI, packages, Runner, and release paths. > - An ordinary PR starts 48 jobs even when its diff affects only a small part of the repository. > - Repeated pushes also repeat successful tests with unchanged inputs. > - This PR selects affected suites before runner allocation and caches deterministic task results with Turbo. > - Required checks validate the plan. Unknown inputs and failed comparisons keep full verification. > - The benefit is fewer runner jobs and less repeated work, with full post-merge verification retained. ## Linked Issues or Issue Description **What existing behavior does this improve?** The GitHub Actions PR workflow repeats unrelated suites for each push. **Subsystem affected** Cross-cutting CI routing, test execution, and required-check aggregation. **Current behavior** An ordinary UI component change starts the complete 48-job PR workflow. Existing dependency and compiler caches do not reuse successful test results. **Proposed behavior** A trusted planner reads the tested merge diff and task inputs. A UI component change such as `ui/src/components/Sidebar.tsx` selects 18 jobs, including build and browser coverage. Successful general, serialized, and standalone Rust tasks can reuse unchanged results across pushes in the same PR. **Reason and benefit** Reduce unrelated jobs and repeated test execution. The 48-to-18 comparison measures planned job count. Remote wall-clock savings have not yet been measured. **Breaking changes** No application API changes. Required `verify` and `e2e` names remain stable. `verify` also checks serialized suites and the canary dry run. ## What Changed - Added explicit task inputs and a pinned Turbo version. An execution adapter calls existing package scripts without adding script aliases. - Added trusted diff planning and dynamic matrices with complete fallback matrices. - Added PR-scoped task result caches. External-state and revision-stamped checks remain uncached. - Added and created the `full-ci` override label, plus repository modes `enabled`, `shadow`, and `full`. Forced and shadow runs bypass result caches. - Added candidate-workflow qualification, planner and aggregate tests, and a real Turbo cache fixture. - Kept the repository-wide CLI guidance guard in the uncached typecheck lane so UI-only diffs retain this security check. - Documented rollout controls, input maintenance, and local inspection commands. ## Verification - Passed 537 focused CI policy, routing, matrix, cache, and shard tests, plus 146 existing release workflow tests. - Passed actionlint for all four changed or added workflows and `git diff --check`. - Passed `pnpm -r typecheck` and `pnpm build` on the implementation base. - Checked real repository task plans for existing general-test and native Rust commands. A real server partition passed, then restored from cache: Turbo task time was 20.022 seconds for execution and 151 ms for the cache hit, excluding wrapper/setup overhead. - The real Turbo fixture exercises the execution adapter and proves source, argument, toolchain-scope, and environment invalidation, result restoration, and retry of failed tasks. - Stopped the serial local `pnpm test:run` after over 30 minutes in favor of the complete parallel GitHub suite. A local parallel database-lane attempt had PostgreSQL startup errors and timeouts. The direct migration fixture passed, then all 273 database tests passed with one local worker. This recovery does not establish the exact cause of the earlier startup failures. The full GitHub PR suite and candidate qualification passed on commit `0bf07ae071c9fe7903edc604d44b5a175727fed0`. Greptile is 5/5 with no unresolved review threads. - All 40 CLI guidance safety tests passed with the exact uncached workflow command. - Verified that master rulesets require only `ci / verify` and `ci / e2e`; both names remain unchanged. - Trusted workflows load from master. The added GitHub-hosted qualification workflow exercises this candidate planner before merge. ## Risks - Changes to global inputs, including task definitions and execution helpers, force execution and disable cache restore/save for the full PR diff. Missing a cross-package filesystem read could omit a suite. Inputs include known server reads of UI files and shared tests of UI assets. Unknown paths, malformed or truncated diffs, and definition download failures select all suites. - Task result reuse requires complete inputs. Keys include Git source files, scripts, resolved dependencies, environment, toolchain, checkout path, and unclassified source files. Generated build artifacts are excluded from input hashes. Registry smoke, browser, and revision-stamped tasks remain uncached. - Maintainers must update task inputs when suites start reading new paths. `shadow` reports selection while running all work. `full` and `full-ci` disable selection and result reuse. - The implementation becomes active after merge. Full post-merge verification remains unchanged. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, repository inspection, tool use, and code execution. The session does not expose a verifiable exact runtime model ID or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
54cc522e86 |
feat(runner): add gated native Antigravity ACP execution (#18188)
## Thinking Path > - Paperclip manages agents and their work. > - Native Runner must bind each provider turn to the selected task, account, model, and instructions. > - Antigravity needs native ACP admission and physical process ownership across continuation. > - A saved conversation must not reuse an old process owner or silently switch its effective model. > - This pull request adds the native candidate behind exact operator admission. > - The default release gate remains closed while qualification is pending. ## Linked Issues or Issue Description Builds on merged #18187. Product account setup and the candidate producer are in dependent #18189. The repeatable browser journey is in #18190. Refs: #12503, #14938; those proposals use legacy CLI adapters. **Agent or provider** Google Antigravity ACP 1.4.3. **Why this adapter is useful** Users need to set up an Antigravity agent and run real work through their managed Google subscription. **How the agent is invoked** Native Paperclip Runner uses the pinned Antigravity ACP runtime. Exact operator admission is required while qualification remains pending. ## What Changed - Add native Antigravity ACP model and instruction admission, selected-account homes, and fenced refresh ownership. - Carry exact candidate authority from the control plane through Runner, Rust, the sidecar, and each current run. Clamp daemon lifetime to the absolute permission expiry, including replacement owners. - Preserve a rotated OAuth refresh token before later identity verification can fail, including failed or cancelled startup admission. Make private lease-close retries safe. - Bound native probe retirement and retry the exact original owner; do not delete credential homes before physical proof. - Preserve Gemini API-key recovery when shared Google subscription capability is added. - Resume the same tagged native conversation with fresh per-turn process ownership and physical native retirement. - Include native manifests in isolated Rust image stages, release cache keys, and dev rebuild detection. The workflow changes only extend existing cache input lists. - Keep current master provider behavior and append Grok 13, Pi 35, and Copilot 84 input identities. Preserve prior identities, including saved Copilot 82 and 83 runs. ## Verification - Runner TypeScript main checks, profile generators, Rust formatting, and offline workspace checks pass. The native backend integration target passes 28 tests. - The complete transport and cleanup group passes 387 tests without unhandled rejections. These checks use local scripted children and actual process and port ownership. - Saved-profile decoding, native-execution parsing, installation, and selected-credential controls pass 208 focused tests. All 109 pre-existing profile identity files remain byte-for-byte unchanged. Copilot, Grok, and shared TypeScript/Rust generation checks pass. - The final recovery fixture passes all seven cases. After the master rebase, the combined recovery and owned-Node-runtime group passes all 10 cases. Each expiry case sets one 45-second absolute grant before startup. It includes the deliberate six-second cold delay, four existing eight-second command budgets, and active-turn headroom. The test proves an active turn before expiry and then checks the daemon, owned process group, child processes, and all three native port fences. The 60-second test deadline does not renew provider authority. - The consolidated model-discovery and pending-release group passes 22 tests. Google credential compatibility and auth recovery pass 25 tests, including existing Gemini API-key recovery. The full product settings test group remains intact in #18189. - Earlier failures remain in their original CI runs and local records. They exposed the Copilot 82 decoder gap, an expired five-second fixture bootstrap ticket, and a fixture grant consumed before activation on loaded CI. Each repair has a separate result. Production ticket limits and admission behavior are unchanged by the final test commits. - An unrelated decision-expiry check failed in [the earlier CI run](https://github.com/paperclipai/paperclip/actions/runs/40941986698/job/123688668997). Its owning code is unchanged from the base. The unchanged suite passes 20 tests on isolated PostgreSQL. A controlled database rollback reproduces the earlier wake: clean startup schedules the 120-second decision expiry, while rollback reconciliation leaves a 30-second continuation-coalescing wake. The decision stays open, with no wake delivery or error. The fresh final-head server shard passes; the failed result is preserved. - Final native commit `ec6bc7e4d77f5ae0cc4946dbba6c662e95a87f5b` passed `ci / verify` and `ci / e2e`. Fresh Greptile and Paperclip reviews both score 5/5, with no open findings. Antigravity retains the recorded native `.5` closure and exact low-model profile. This PR does not claim global or cross-platform qualification. ## Risks - The new transport carries private managed credentials. Exact ownership, account generation, model, instructions, and native retirement are fail-closed. - Only local macOS ARM64 has a verified complete Antigravity runtime. Default Antigravity task execution remains disabled without current trusted operator admission. The Google server sign-in lane and production candidate producer ship in #18189; this native slice does not expose the subscription catalog choice. - Shared launch inputs change provider identities. New identities are append-only; existing recorded identities remain available. ## Model Used OpenAI Codex (GPT-6). The host did not expose the exact model revision or context limit. The model used tools for code edits, tests, and review. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandPaperclip
|
27be859044 |
Record runtime activity under Runner authority (#18050)
## Thinking Path > - Paperclip manages agents and tasks. > - Runner owns each authenticated run and its provider process. > - Runtime clients need to report successful activity in that run. > - A retained process can outlive a run, so an operation needs current authority. > - This PR captures that authority before the operation and records activity afterward. > - Core can consume the event through the existing run stream. ## Linked Issues or Issue Description Builds on merged #18129 and targets master. The dependent consumer is #18131. Test stability fixes were separated into merged #18128. Earlier combined-head verification remains in [the archive](https://github.com/paperclipai/paperclip/pull/18050#issuecomment-6489369676). **Subsystem affected** Runner activity recording and provider command environments. **Problem or motivation** A retained provider process can complete an operation after its original run ends. Runtime clients need a host-attributed event and a helper that works inside the provider sandbox. **Proposed solution** Capture opaque Runner context before use. Publish a structured notification after success. Runner checks current authority and records `runtime.activity.recorded`. ## What Changed - Add `notification-context` and `record-activity CONTEXT NAMESPACE ACTIVITY_KEY`. Rotate context on warm attachment and process recovery. Accept no caller run ID. - Use one private named pipe. Reuse Runner's bounded JSON-line reader. Serialize concurrent publishers with a read-only lock. A blocking reader queues at most 128 records; the existing control loop handles at most 16 per pass. - Grant Codex access to the exact helper, descriptor, lock and pipe. Preserve other filesystem and network limits. Preserve host discovery values through provider launch boundaries and reject configured-environment overrides. - Update source identities to Pi 34, Copilot 81 and Grok 12. Keep historical identity fixtures and recovery behavior. ## Verification - Targeted release notification and sandbox-policy tests: seven passed. They cover run rotation, recovery, Unicode limits, concurrent writers, malformed and partial records, bounded queues, failed persistence and fixed-path cleanup. - Real CLI integration and pinned Codex 0.160.0 sandbox checks: passed. The sandbox check uses no model, credential or network call. It verifies activity recording while unrelated state reads and parent-directory writes remain denied. - ACPX profile generator check, formatting and diff checks: passed. - Rebased onto master `31b927e30` at head `64d0248a8`. Master used the same profile versions, so the resolution preserves its Pi 33, Copilot 79 and Grok 11 fixtures byte-for-byte and creates new identities. All identity checks pass. Five targeted provider/environment/recovery files pass 168 tests with one existing skip; the Rust Pi admission regression passes. The notification module and Codex sandbox implementation are byte-identical to the tested pipe head `671373fe0`. - Removed the mailbox traversal reported by security review. Its regression replaces the pipe path with a 100-level tree and an unreadable leaf. Shutdown preserves the leaf, performs no traversal and completes promptly, including with a partial record. - An initial targeted test failed because its run-rotation fixture updated config without updating durable state. The corrected fixture passes. Runner also rejects inconsistent run bindings. - Previous head `26abbf62b` passed Runner, build and browser CI. Server shard 11/12 failed in unchanged `attention-service.test.ts`: cleanup/test timeouts were followed by a duplicate `ATQ` fixture key. The final head needs fresh CI and review. - CI on `64d0248a8` completed with 52 successful checks and four intentional skips. Runner Vitest shard 1/2 failed when retained-state copies encountered the live FIFO; the aggregate also failed. Standard review scored 4/5 and found relative pipe paths plus late context publication. Follow-up `5c3f14d32` places IPC in a private canonical temporary directory and publishes context before welcome commands. Its relative-state CLI test, two authenticated reconnect cases, seven notification/policy cases and real Codex sandbox pass. One affected retained-settlement case passes with the matching release binary; the other 18 cases remain for CI. The sandbox denies sibling-directory and lock writes. All build, test and security checks pass on `5c3f14d32`; standard review is 5/5 with no unresolved threads. The requested Apex review is 5/5 on this exact head, with no findings or unresolved threads. - Earlier broad test results in the archive apply to their recorded source revisions. Per the requested workflow, final local checks stay targeted; GitHub CI supplies broad coverage. - After Apex completed, master advanced again and conflicted on Copilot identity. Head `e94a9359d` is rebased on master `8ac5a128e`; it preserves the published Copilot 80 fixture and uses Copilot 81. The notification module, Codex sandbox and durable control-loop code are byte-identical to Apex-reviewed `5c3f14d32`. All 131 targeted installation/compatibility/recovery cases and identity generator checks pass. All 55 checks pass with four intentional skips on the rebased head. Standard review is 5/5 with no unresolved threads. Apex is 5/5 on `e94a9359d`, with zero findings and no unresolved threads. GitHub confirms this head is conflict-free and ready for review. ## Risks Activity remains an optional hint and grants no access. A successful helper exit means publication to the pipe; Runner persists accepted events through its existing durable outbox. Full queues, unavailable helpers or failed writes can drop a hint without failing the operation. Context reads are local and do not wait for the control loop. Publishers have a 500 ms lock deadline and a 2 KiB record limit. IPC lives outside durable state. Idle transport blocks; cleanup removes its fixed files and empty private directory. The host must provide `/usr/bin/mkfifo`. Rebuild matching Runner/provider artifacts for the updated source identities. No dependency, migration, live provider call or deployment is included. ## Model Used OpenAI Codex, based on GPT-6, with reasoning, repository tools and code execution. The exact serving snapshot is not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
2a6d4d1694 |
perf(db): index canonical task activity lookups (#18043)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/177/changes/paperclip/177 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip helps people manage AI agents and their tasks. > - Task lists rank tasks by their latest canonical activity. > - The existing lookup reads every event for each task before pagination. > - Query-only alternatives failed a narrow visibility control or the performance goal. > - This PR adds an index that lets the unchanged lookup stop at the newest qualifying event. > - The result is less history work with the same permissions, ordering, and response. ## Linked Issues or Issue Description Refs: #13803. That broader proposal stores a new activity field with triggers and a backfill. This index was authored independently and uses the existing query. **Subsystem affected** Task-list activity ranking and the database schema. **Current behavior** The list computes `MAX(activity_log.created_at)` for every eligible task. The existing entity index reads 50 events per task across 1,199 candidates in the representative fixture. **Proposed behavior** Add `(company_id, entity_type, entity_id, created_at)`. Keep the production task service and permission helpers unchanged. PostgreSQL reads backward until an event passes the existing exclusions. **Reason and benefit** Fresh paired measurements on the current base improve the full-description scoped SQL from **44.60 to 20.33 ms (2.19×)**. Activity buffer accesses fall from **63,547 to 4,796**, and history rows per probe fall from 50 to one. The complete service median falls from 50.63 to 31.31 ms. Complete positional SQL results and service responses match. ## What Changed - Add generated migration `0414_closed_edwin_jarvis.sql` and the matching Drizzle index. - Use `CREATE INDEX CONCURRENTLY IF NOT EXISTS` through the existing top-level runner and invalid-index retry cleanup. The client change is one whitelist entry. - Preserve all upstream SQL and journal entries; retain snapshots 0410–0414. - Cover full and index projections, canonical ordering, invalid-index recovery, concurrent migrators, and incomplete fixture rejection. - Repair benchmark seeding, apply requested work memory to both connections, and record both settings. Hosted defaults now include six Slack conversations, 642 GitHub conversations, and 53 removed conversations. ## Verification - **All final-head CI gates passed**, including full workspace typechecks, tests, build, Runner checks, and all eight E2E shards. Greptile is **5/5** on `67ca7ae1cf15fc054ef515ffd8205b1fe30a0d55`; no unresolved review threads remain. - Local database/server typechecks and migration numbering/safety checks pass. Three server regression tests and three migration regression suites pass on real PostgreSQL. - The original pre-index work guard fails at eight rows per probe; the indexed guard now checks both the full-description query and index projection. The service oracle covers cleanup/inbox exclusions, raw text IDs, company scope, comments, ties, sort directions, visibility before pagination, empty lists, and one child per parent. - The documented pre-index seed command reproduced the review failure before the fix and now succeeds. A paired run with non-default memory confirms both independent connections use `16MB`. - Fresh proof pins pre-index base `929e840595c8933ca485ac7a2e99413e1e11ca64`, with identical cloned data, unchanged generated SQL and parameters, two warmups, and five alternating samples per arm. All **18 comparison cells** preserve every positional row and complete service response. The 15-case default-memory matrix passes its physical-work, speed, and control bounds. - The one-visible-task control is 6.64 → 7.25 ms SQL; the 20-task control is 3.71 → 3.72 ms. Both pass the declared 25% or 2 ms budget. Full-query memory controls use matching `4MB` or `16MB` settings. - Committed 1,000-row write batches measured 40.09 → 43.17 ms median (+7.7%). Ranges overlap (34.67–71.85 and 34.27–48.05 ms); this is an observed local cost, not a precise throughput guarantee. - The final-numbered migration builds a valid index through the normal runner on the fresh 300,000-row fixture in 0.55 s. Earlier large-fixture proof of the same index SQL used 5,000 tasks and 1.25M activity rows: 2.58 s first apply, concurrent write/read probes 5.16/2.43 ms, interrupted invalid-index cleanup, successful retry, and one receipt. Current migration recovery tests verify the renumbered path. - [The proof document](https://github.com/paperclipai/paperclip/blob/67ca7ae1cf15fc054ef515ffd8205b1fe30a0d55/docs/performance/task-list-activity-index.md) retains the historical baseline, rejected rewrites, and the earlier matrix invalidated by severe host contention. Those historical timings are not substituted for the fresh measurements above. ## Risks The migration is an exception after measured query-only alternatives failed. The index adds storage and work to every activity write: about 30 MiB for 300,000 rows and 123 MiB for 1.25M rows. Concurrent creation avoids the writer blocking observed with an ordinary index build, but long transactions can delay it. Local timing does not establish Neon build duration. The query still ranks all eligible tasks, and excluded newer events still need to be skipped. These are local SQL/service measurements; deployed HTTP latency remains unverified. No hosted DDL or deployment was performed. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, code execution, and tool use. The environment does not expose the exact model identifier or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either linked existing issues or described the issue using the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links - [x] My branch name describes the change and contains no internal issue id - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandPaperclip
|
aa75ec9d56 |
Replay Claude native options on ACP reconnect (#18135)
## Thinking Path > - Paperclip manages agents and their task execution. > - Runner uses ACPX to connect to Claude sessions. > - Claude receives native options when ACPX creates a session. > - ACPX omits those options when it loads or resumes the same session. > - A reconnect can lose allowed tools, instructions, and settings sources. > - This pull request replays the existing options through the existing Claude metadata contract. > - Resumed tasks keep the configuration admitted for their session. ## Linked Issues or Issue Description **What happened?** A Claude task can complete its first turn and then fail on a warm follow-up. The provider refuses an internal completion tool because `session/load` omits the `allowedTools` options sent with `session/new`. The same omission exists on `session/resume`. **Expected behavior** Creation, load, and resume must supply the same caller-selected native session options. The existing policy for native user settings must stay unchanged. **Steps to reproduce** Create a persistent Claude ACP session with an explicit allowed-tool list and appended instructions. Complete a turn, close its connection, and reconnect to the saved session. Inspect the ACP request. Before this change, the reconnect has no `_meta` field. **Paperclip version or commit** Reproduced on Runner source `e8b78cce5bb557b086bee9ca930eeaab1d09b244`. The offline protocol regression also fails on the current base `80a2b9284bf189e93ca4313085814548f48b8dc8`. Related: #18089 added prepared harness storage. #11798 concerns the separate legacy adapter and does not fix Runner reconnect metadata. ## Upstream status and backport scope ACPX [0.21.1](https://github.com/openclaw/acpx/releases/tag/v0.21.1) already replays session metadata on load and resume in its [released client implementation](https://github.com/openclaw/acpx/blob/v0.21.1/src/acp/client.ts). This PR backports that behavior to the currently qualified 0.13.1 integration. The released 0.21.1 runtime does not expose equivalents for all hooks used by Runner's existing patch: verified executable spawning, synchronous protocol admission guards, provider extension dispatch, and response-delivery receipts. Upgrading requires a compatibility pass that preserves these guarantees and removes local changes already covered upstream. This backport does not add a new Paperclip-specific dependency contract. ## What Changed - Replay existing Claude native options on `session/load` and `session/resume` through the same metadata builder used at creation. Other providers keep their existing reconnect requests. - Test both reconnect methods against the installed ACPX manager and a local ACP wire fixture. Cover allowed tools, model, turn limit, instructions, and both existing settings-source modes. - Update exact source attestations and generated profile declarations required by the shared ACPX patch. Keep historical fixtures unchanged. - Document the reconnect contract. No new home mode, plugin policy, or provider permission rule is added. ## Verification - Confirmed the new load and resume tests fail before the patch because metadata is absent. - Applied the canonical patch to pristine package bytes and confirmed it matches the tested runtime. - Targeted installed-package model-selection and reconnect suite plus changed Grok identity tests: 22 passed. - Rechecked the four new reconnect cases against the cleanly applied patch, with explicit expected values for all supplied options. - Three targeted current-profile and revision-boundary assertions pass locally. Updated the native Pi test digest; native compilation and execution remain in CI. - Maintained profile generators pass their consistency checks. - GitHub CI is green, including typecheck, native Runner checks, package contracts, packaged installation, build, and server/browser suites. One unrelated server database-cleanup timeout passed on the bounded failed-job rerun. No comprehensive local suite was run. Live post-deployment acceptance remains pending. ## Risks Claude reconnects now receive the same native settings sources and options as creation. This corrects configuration loss without changing caller policy. The shared dependency patch requires new source identities for other attested profiles; their reconnect behavior is unchanged. No database migration is required. ## Model Used OpenAI Codex, GPT-6 family, with code and tool use. Exact deployment ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Nicky LeachandPaperclip
|
c57af3c564 |
Reuse prepared Runner storage and installed skill selections (#18089)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/207/changes/paperclip/207 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip manages AI agents and their work. > - The native Runner prepares selected skills, instructions, and provider storage. > - Repeated preparation imports the company skill inventory and rebuilds Claude storage. > - Execution environments can already maintain prepared resources. > - This pull request lets them reuse account storage while Core keeps execution policy. > - New retry runs resolve current skills and keep the failed attempt's history. ## Linked Issues or Issue Description **What existing behavior does this improve?** Native Runner preparation and retry skill selection. **Current behavior** Explicit installed selections still trigger company inventory reconciliation. Plugin preparation receives account identity only on warm transitions. Claude's isolated mode requires a session-specific configuration directory. **Proposed behavior** Skip inventory imports when explicit selections are installed, while retaining current authorization and version resolution. Send account identity on initial preparation. Allow an execution environment to provide persistent Claude storage without enabling preconfigured permission behavior. A new retry resolves current skill assignments. **Reason and benefit** Remove repeated preparation work without adding environment-specific policy to Core. Related: #13984 describes a similar storage-lifetime problem in the separate claude_local adapter; this PR does not change that adapter. ## What Changed - Derive Claude’s configuration path from the single authoritative prepared-storage input; separately supplied CLAUDE_CONFIG_DIR values are overridden. - Add an explicit prepared Claude storage binding, forward it to the sidecar, and retain recovery and provider lifetime fencing. - Refresh Core-owned Claude settings after acquiring the lifetime lease. Reuse environment-prepared skills. - Reuse installed explicit skill selections and preserve missing-selection initialization and current access checks. - Reserve the prepared-home marker for environment provisioning; task overrides are rejected and configured bindings are stripped before native launch. - Clear inherited skill selection snapshots on new retries. Keep admitted execution recovery and failed attempt metadata. ## Verification - Current single-input follow-up `8dbc2f6ec`: 154 focused Runner checks pass with one existing skip; four Grok identity checks and four Pi packaged-probe checks pass. Shared immutable source identities advance to Grok10/Pi32/Copilot74. All56 current-head CI checks pass (four intentionally skipped); standard and enhanced reviews are both5/5 with no unresolved threads. - Focused Runner sandbox tests: 29 passed. Focused environment tests: 16 passed. - company-skills-empty-runtime: 10 passed. - Plugin Runner and runtime selection database tests were attempted. Local execution is blocked because embedded PostgreSQL cannot load libzstd.1.dylib; these checks remain for CI. - New database regressions cover installed-selection reuse, initial account preparation, and retry snapshot removal. - Prepared-home trust-boundary regressions: 4 targeted checks passed (task schema, all env binding scopes, native executor). - Focused Pi public-install probe checks: 4 passed. The build smoke probe now asserts the current Pi 32 profile and exact digest. - CI found and fixed retry-context TypeScript inference, stale Pi identity fixtures, Docker source identity inputs, and the skill-observability test’s old assumption that every selection imports inventory. Cold and reused selection coverage remain. - CI chat shard 2 initially failed three GitHub/Discord idle-database assertions. An unchanged-head rerun passed; the original cause remains unexplained. - Runtime head `f5b53679a` passes all 56 current CI checks (four jobs intentionally skipped), including the full build/public-install probe, TypeScript, Rust, Runner and server matrices. Greptile is 5/5 with no unresolved review threads. - The requested enhanced review scored 5/5 on `f5b53679a`, with one nonblocking documentation request. Documentation follow-up `1ac5d26a8` adds only the documented existing ownership/0700 requirement; the thread is resolved and all 56 current-head checks pass (four intentionally skipped). The subsequent single-input runtime follow-up8dbc2f6ec repeated the complete CI → standard5/5 → enhanced5/5 cycle successfully. - Broader verification is delegated to GitHub CI as requested. The initial policy check rejected a generated lockfile repair; that change was removed because Core CI already resolves stale importer entries. ## Risks Prepared storage must be provisioned by the execution environment before launch. It is not credential isolation. Core replaces managed Claude settings under the existing provider lease. No new per-run harness homes are created by this option. New retries may select a newer skill version; recovery of an already admitted provider execution keeps its saved input. ## Model Used OpenAI Codex, GPT-6 family, with reasoning and tool use. The exact model revision is not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked related context above - [x] I have described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links - [x] My branch name describes the change - [x] I have run focused local tests; database and full verification remain for CI as explained above - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation - [x] I have considered and documented risks - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
52b86c2032 |
perf: avoid repeated run-context reads for queued comments (#18021)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Execution recovery shows how many saved user messages remain unconsumed. > - This read must preserve the rule that consumed input cannot be replayed. > - The current query reads a large run context again for each saved comment. > - PostgreSQL already scans the run relation once, but repeats the JSON reads. > - This pull request projects the two wake-comment fields before the repeated comparison. > - The query uses less database work and returns the same blocker state. ## Linked Issues or Issue Description **What happened?** The eight-comment consumption query averaged 131.14 ms across 99 calls in a recent hosted sample. Each call returned all eight comments. The query used about 8,002 shared buffer hits per call. It used no JIT compilation. The local plan showed repeated TOAST reads in the anti-join, despite a single run scan. **Expected behavior** Count saved, undeleted user comments without reading each large context again for each comment. Keep exact JSON issue bindings, JSON containment, company scope, and cancellation rules. Consumed input must remain excluded. **Steps to reproduce** Run the committed benchmark against disposable PostgreSQL with repository migrations. Pass the pinned baseline below. Seed forty task runs with 72 KiB random contexts, eight unconsumed comments, and 1,400 unrelated runs. Compare complete SQL rows and the full execution blocker result. **Paperclip version or commit** Pre-fix baseline: `8279c990080b15abdcd9981f5793290fab37f9ad`. This branch is based on current master. The intervening master changes do not change the measured service or its SQL helpers. **Deployment mode** The report came from hosted PostgreSQL 17.11. The local proof uses disposable PostgreSQL 17.11 on ARM64 Linux. Related correctness work: #16528, #16664, and #18012. This query change does not replace those fixes or resolve the replay report in #17782. The duplicate search found no open performance fix for this query. ## What Changed - Select only `wakeCommentId` as text and `wakeCommentIds` as JSON before the anti-join. - Use `OFFSET 0` as a PostgreSQL projection boundary. Keep lazy first-match behavior and the existing company/JSON-issue index. - Preserve the original singular text comparison and JSON containment expression. Keep cancelled-before-start, deleted-comment, user-author, and company filters unchanged. - Add real PostgreSQL regression tests and a reproducible paired benchmark. Compare every positional SQL column and the entire service result. - Record the rejected scan-count hypothesis, slower controls, full samples, and acceptance limits in the [proof document](doc/plans/2026-10-28-queued-comment-consumption-query.md). ## Verification - Red before the source change: the eight-comment/300-run fixture used 35,711 shared buffer accesses and failed the selected 10,600-work budget. Five semantic cases passed. The earlier single-scan guard passed the original code and is not counted as red. - Green: six new real PostgreSQL cases and 85 existing queued-comment route cases pass together: **91/91**. Coverage includes cancelled-before-start, started cancellation, no replay, conflicting native IDs, noncanonical JSON IDs, cross-company data, deleted/agent comments, and scalar/object/nested-array JSON. - `pnpm -r typecheck` and `pnpm build`: passed. `pnpm test:run` was attempted and stopped after 16 native PostgreSQL startup failures in an unchanged executor suite; 924 other cases in that file passed. The full local attempt is incomplete and is not counted as a pass. Those same 16 cases then passed with Docker PostgreSQL 17.11. Full Linux CI passed on the final head. - A standalone empty native `initdb` fails with `semget: No space left on device` from the macOS semaphore limit. Native PostgreSQL startup is unsupported on this host. Its focused attempt skipped six tests. The 91-case run used a private Docker PostgreSQL constructor substitute. Repository database creation, migrations, clients, and cleanup were retained. No test infrastructure change is committed. - Final head `d2921d9220e22624f85f7057a577a6e2e98dc097`: all 58 check contexts are terminal and clean (55 success, two expected skips, one neutral). The full [PR workflow](https://github.com/paperclipai/paperclip/actions/runs/40699607944) passed. Greptile reviewed this exact head at 5/5. Paperclip Review originally requested the deep first-consumed control; it now passes and that thread is resolved. Its latest automated review attempt stopped before completion and reported neutral, so no fresh Paperclip 5/5 is claimed. Two warmups and five alternating samples per arm, with capture, equality checks, and plans outside the timers: | Local fixture | SQL median before → after | Full blocker median before → after | Shared buffer accesses | | --- | ---: | ---: | ---: | | Representative: 40 runs, 72 KiB, 8 unconsumed comments | **11.47 → 3.82 ms (3.00×)** | **29.48 → 18.85 ms** | **7,686 → 966** | | Same fixture, JIT disabled | 10.96 → 3.63 ms | 32.18 → 24.75 ms | 7,686 → 966 | | 11 runs, 72 KiB | 8.35 → 5.05 ms | 37.60 → 35.13 ms | 2,116 → 268 | | 8 comments consumed by the first wide run | 1.60 → 1.28 ms | 13.90 → 16.59 ms | 196 → 28 | | 1,400-run stress, above observed task depth | 172.82 → 44.27 ms | 212.60 → 73.94 ms | 175,089 → 37,889 | | 1,400 runs, first wide run consumes all 8 | 1.54 → 1.99 ms | 19.00 → 18.23 ms | 122 → 24 | Representative SQL ranges: 10.36–12.33 ms before and 2.83–4.30 ms after. Service ranges: 27.34–33.22 ms before and 18.58–21.81 ms after. All 24 final fixture/JIT cells return identical complete rows and service objects. The follow-up widens the consuming run to the same payload size as history and supersedes the initial thin-row consuming controls. All final-matrix benchmark budgets pass. The deterministic buffer-work budget was selected before the production change. The timing criteria were selected after the initial exploratory projection trial and before the final quiet matrices. Small and empty-history service medians rose by 5.03 ms and 3.90 ms in the JIT-on matrix. Small-case ranges overlap: 18.49–36.48 ms before and 17.29–41.89 ms after. These differences do not establish a causal service regression. The JIT-off small and empty-history controls improved slightly. The unchanged no-comment path runs no changed SQL. The one-comment control is neutral. The widened immediately consumed service median rose 2.69 ms (19 percent), inside the selected 25 percent or 2 ms service budget. Its ranges overlap: 13.34–27.72 ms before and 13.18–18.10 ms after. This accepts a possible bounded small-path cost; it does not claim every path gets faster. All controls and full distributions remain in the proof. The reviewer-requested deep first-consumed plan reads exactly **one** underlying run row in each arm. Both materializations return one row, revisited eight times. The remaining history is not read eagerly. JIT-on SQL ranges overlap at 0.86–2.79 ms before and 0.77–4.38 ms after; this proves early exit rather than a latency gain. With JIT disabled, SQL is 1.08 → 1.07 ms and service time is 11.36 → 11.53 ms, with the same 122 → 24 buffer accesses. Production source is unchanged by this proof update. Reproduce: ```sh QUEUED_COMMENT_BENCHMARK_DATABASE_URL="$DISPOSABLE_QUEUED_COMMENT_DATABASE_URL" \ pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-queued-comment-consumption.ts \ --baseline-ref=8279c990080b15abdcd9981f5793290fab37f9ad --require-budget ``` Use an isolated `queued_comments_` database. Repeat with `--jit-off`. Without the URL, the script creates disposable embedded PostgreSQL. The benchmark rejects changed runtime helpers and records source hashes, settings, SQL/parameters, complete results, samples, and plans. ## Risks - PostgreSQL planner changes can alter the materialization strategy. The regression guard checks repeated context work on the supplied fixture. - Small local service medians are noisy. Their ranges and slower medians remain visible. A faster local query does not establish the same hosted request saving. - JSON authority and no-replay rules are sensitive. The query keeps the original predicates and has real database tests for non-array and conflicting-ID cases. - There is no migration, index, cache, history cutoff, extra round trip, deployment, or hosted write. ## Model Used OpenAI Codex, GPT-6 family, with reasoning and tool use. The exact model ID and context window are not exposed by this session. They are not inferred from a model alias. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
MichielMAnalyticsandClaude Opus 5.5
|
644937c931 |
feat(sandbox-providers): add boxd sandbox provider (#17284)
<p align="center"> <img src="https://raw.githubusercontent.com/MichielMAnalytics/paperclip/08f1cdda04443ed10fc43e7097ae0b2d14b5cc19/boxd-x-paperclip.png" alt="boxd × Paperclip" width="700"> </p> ## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - Agent work runs in sandboxes that provider plugins supply > - An operator selects a provider for each environment, and a reused lease keeps one sandbox for an issue across heartbeats > - The time to resume a reused sandbox delays every heartbeat on that issue (see #14684) > - boxd runs each sandbox as an isolated microVM that suspends and resumes in milliseconds, with disk, memory and processes kept > - This pull request adds a boxd sandbox provider plugin > - The benefit is that operators get one more provider, with fast resume for reused leases ## Linked Issues or Issue Description Refs #14684. That pull request keys the remote session identity on the provider lease ID. This provider keeps the provider lease ID equal to the boxd machine ID, so the ID stays the same across suspend and resume. No issue exists for boxd. The request is described below with the fields of `adapter_request.yml`. **Agent or provider** boxd (https://boxd.sh). boxd runs each sandbox as a KVM microVM. The plugin uses the public TypeScript SDK, `@boxd-sh/sdk` 0.2.20 or later. **Why this adapter is useful** boxd suspends a machine in about 17 ms and resumes it in about 0.2 s, with the disk, the memory and the running processes kept. This fits `reuseLease`: the next heartbeat on an issue continues on the same machine. Each machine is created with `isolated: true`. An isolated machine has no metadata endpoint, no in-VM `boxd` CLI and no boxd integrations, so sandbox code cannot reach the provider management interface. **How the agent is invoked** Build and install the plugin by local path, as the package README describes. Open Instance Settings, then Environments. Select the `boxd` provider and supply a boxd API key. Agents that use the environment run their adapter commands in a boxd machine. **Are you willing to implement it?** Yes. This pull request is the implementation. ## What Changed - Adds the `boxd` sandbox provider under `packages/plugins/sandbox-providers/boxd`. Its structure follows the E2B and exe.dev plugins. The plugin has no code comments, to follow `CONTRIBUTING.md`. - Implements the validate, probe, acquire, resume, release, stop, destroy, inspect, realize-workspace and execute hooks. The manifest declares `sandboxCapabilities: { reusableLeases: true }`. - With `reuseLease`, release suspends the machine and resume wakes it. A suspended, hibernated or stopped machine is woken by the matching boxd call. A release that cancels active work stops the machine and keeps its disk. - Release and destroy return a `destroyed` receipt only after boxd no longer lists the machine. Stop returns a `stopped` receipt only after boxd reports the `stopped` state. - When the host sets `requestedExpiresAt`, the plugin sets a boxd delete timer at or before that time and returns `expiresAt` on the lease. The timer is set after the create call returns, so a failure while setting it cannot lose the machine ID. - Commands run through `sh` with quoted arguments and per-command environment. Stdin is staged with boxd's file API in a temporary file that only the machine user can read, and removed afterwards; the workspace directory is created the same way. A machine whose boxd agent predates that API, such as one restored from an older snapshot, falls back to the older upload and `mkdir` calls. A command timeout ends the remote process group and returns `timedOut: true`. A command on a stopped machine starts the machine and runs one more time. - A failed acquisition deletes its machine. If boxd cannot confirm the deletion, the plugin throws `PluginEnvironmentCreationCleanupError` with the machine ID, so the host records it and retries the teardown; the destroy hook checks the account and the machine labels before it deletes a failed creation. A probe whose machine cannot be deleted reports a failure, and probe machines carry a 15-minute boxd delete timer. - The plugin keeps one SDK client for each API key. This prevents a session token exchange for each command. - Machines carry the labels `paperclip-provider`, `paperclip-environment` and `paperclip-run`, so an operator can find them with `boxd machine list --label paperclip-provider=boxd`. - Registers the optional bundled catalog key `boxd` and the trusted `BOXD_API_KEY` host fallback, with tests that mirror the CreateOS cases. - Adds the package to the release manifest with `publishFromCi: false` until a maintainer bootstraps the first npm publish. - Adds unit tests with a mocked SDK (40 tests) and opt-in live tests (`BOXD_LIVE_TEST=1`, 9 tests) that run every hook against boxd. - Adds a section to `doc/plugins/PLUGIN_AUTHORING_GUIDE.md` and adds boxd to the README provider list. ## Verification Provider package (`packages/plugins/sandbox-providers/boxd`): - `pnpm install --ignore-workspace --no-lockfile && pnpm typecheck`: passes. - `pnpm test`: 40 passed, 9 live tests skipped. - `BOXD_LIVE_TEST=1 BOXD_API_KEY=... pnpm test`: 49 passed against production boxd on 2026-10-28 with `@boxd-sh/sdk` 0.2.20. The live tests cover: - validation, and a probe at 2 vCPU that deletes its machine (a bad key returns a failed probe); - the labels, the isolation and the org of an acquired machine; - stdin, env, argument quoting, the exit code and stderr, and the removal of the staged stdin file; - an empty stdin, and a staged stdin file that only the machine user can read (mode 600); - a 5 MiB round trip through the base64-over-exec path (the SHA-256 matches); - a command timeout, which returns `timedOut` and ends the remote processes; - suspend and resume (a file and a running process stay), stop and resume (the file stays), and inspect in each state; - a command on a stopped machine, which starts the machine and runs; - resume and stop of a hibernated machine, stop of a suspended machine, and release with `cancelActiveWork` and with `stop_and_retain`; - release without reuse, which deletes the machine; - a `requestedExpiresAt` 45 s away, which boxd enforces by deleting the machine; - a run restored from a snapshot, and a fresh machine at 2 vCPU / 8 GiB; - resume after destroy, which returns an expired lease, and a final check that no test machines remain. - Measured times: probe 1.5 s, acquire 1.4 s, release with suspend 17 ms, resume from suspend 0.13 s, resume from hibernate 0.18 s, stop 14 ms, resume from stop 3.2 s, acquire from snapshot 1.6 s, confirmed destroy 30 ms. - On a machine started before boxd's file API (the test host), the plugin's hooks fell back to the older calls: the workspace was created, stdin was staged (mode 644, the upload path), an empty stdin worked, and no staged file was left. Host: - `npx vitest run src/__tests__/bundled-plugins.test.ts src/__tests__/plugin-database.test.ts` in `server`: 73 passed, 6 of them new (re-run on 2026-10-28). - `npx vitest run src/__tests__/cli-invocation-safety.test.ts` in `server`: 40 passed. The README install commands now use the direct-exec CLI form. - `npx tsc --noEmit` in `server`: passes. - `node --test scripts/release-package-map.test.mjs scripts/check-release-package-bootstrap.test.mjs`: 18 passed. Full stack, on a Linux host with `pnpm dev:once` in `local_trusted` mode: - I installed the plugin with `POST /api/plugins/install` and a local path, then created boxd environments with a secret reference for the API key. - Ephemeral environment: the environment probe passed. A `claude_local` agent (`claude-sonnet-5-5`, legacy runner) ran a task in a boxd machine. It reported Ubuntu 24.04.5 and the workspace `/home/boxd/paperclip-workspace`, and it marked the task done. The lease was released and the machine was deleted. - Reused-lease environment: the first run wrote a file and started a background loop. After release, the machine was in the `suspended` state. A comment with `reopen: true` started a second run. That run used the same machine, so the provider lease ID was the same on both lease rows. The file was there and the background loop was still running. - Validation: a config with `vcpu: 3`, and a config with both `vcpu` and `snapshot`, return 422 with the plugin error. The unsaved-config probe (`/environments/probe-config`) passed in 2.0 s and deleted its machine. A saved probe with `vcpu: 4` created a 4 vCPU / 16 GiB isolated machine. - Cancel: cancelling the first run on a reused lease marked the lease `expired`, and the host destroyed the machine through the plugin about 12 s later. - Delete: deleting an environment that holds a reused lease returns 409. With `?destroyReusableSandboxLeases=true` it returns 200, and the host destroys the suspended machine through the plugin. - I re-ran this flow on 2026-10-26 and again on 2026-10-28 with `@boxd-sh/sdk` 0.2.20, from a script, with each step checked. On 2026-10-28 the ephemeral run posted its proof and both of its machines were deleted. The reused-lease follow-up ran on the same machine (both lease rows have the same provider lease ID), and found the file and a background loop that had kept running. The validation errors returned 422, the unsaved-config probe passed in 1.8 s, and deleting the environments left no machines. Every check passed except the host gap below. - After the runs, `boxd machine list` showed no machines with the test prefix. Not run: `pnpm -r typecheck`, `pnpm test:run` and `pnpm build` at the root. The test host has no Rust toolchain for the runner binary. On that host, `scripts/check-forbidden-tokens.mjs` adds the local username (`boxd`) as a token. It then flags every `boxd` substring, and it also fails on `master` there (`InboxDismissal`). A CI runner uses a different username. ## Risks - Host gap, not in this plugin: when a run that **resumed** a reused lease is cancelled, the host marks the run `cancelled` but does not stop the agent in the sandbox. The adapter keeps polling the file bridge, and the host calls no release, stop or destroy hook. The lease stays `active` until the environment is deleted. A plugin-side trace showed only bridge-polling `execute` calls after the cancel. boxd is the only provider that declares `reusableLeases`, so this path may not have been exercised before. A cancel on the first run of a lease works as expected. - Low risk to existing behavior. The plugin is opt-in and is not installed by default. The host changes add one catalog entry, one credential passthrough entry and their tests. - boxd machine creation has no idempotency key. If a create response is lost, a machine can stay behind. The labels above help an operator find it. - A suspended machine keeps its memory image until boxd hibernates it after a period of idle time. An operator who keeps many reused leases pays for that storage. - boxd stops a non-PTY command that has no output for 30 minutes. Agent CLIs write output during a run, so this does not affect normal runs. - boxd has one region (EU), amd64 machines only, and no GPUs. Machine sizes are 1/4 GB, 2/8 GB, 4/16 GB and 8/32 GB. - The new `package.json` needs code-owner approval. ## Model Used - Claude Opus 5.5 (`claude-opus-5-5`) through Claude Code, with tool use and shell execution on a boxd machine. The session does not expose the context window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (focused checks; the limits are listed under Verification) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01HNWtBYxLByQmmNKjuaou1P --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
DottaandPaperclip
|
d8bd8fe9fb |
feat(evals): browse E2E suites across targets and models (#18027)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - E2E tests show whether people can complete real app and agent journeys. > - Operators need to choose a suite and inspect its evidence across deployments and models. > - The history page listed campaigns without a suite browser or clear coverage filters. > - QA smoke also appeared separate from the other retained E2E measurements. > - This pull request puts suites, deployment targets, and named selections in one history browser. > - The benefit is that a passing subset stays distinct from a full suite run, with direct links to evidence. ## Linked Issues or Issue Description Related: #18009 added clear per-test details and links to the tested application. This PR extends the shared E2E history browser. **What happened?** The public directory called full app testing “Product E2E Evals” and made users choose campaigns before they could find a suite. The history list could not filter independently by tested app, agent execution, model, or run selection. A recent passing subset could be mistaken for a full suite result. **Expected behavior** Browse current and retained suites in one E2E category. Keep QA smoke in normal campaign history. Show the latest matching run and latest full suite run. Preserve the recorded definition and denominator when filters change. **Steps to reproduce** 1. Open the E2E history directory. 2. Find a suite, then compare a narrow passing run with a full run across execution environments and models. 3. Find the QA smoke measurement and open its retained screenshots. **Paperclip version or commit** The previous report UI was at `8279c99008`. This branch is rebased on current master. **Deployment mode** Static E2E history and campaign reports generated from retained measurements. ## What Changed - Call full-product testing “E2E tests.” Keep Runner protocol evals separate because they use a mock control plane. - Add a searchable suite sidebar that includes current suites and suites found only in retained history. - Add correlated suite, coverage, deployment target, execution, provider, model, profile, runtime, result, date, and named-selection filters. Keep filters in the URL for reload and back navigation. Sidebar suite links clear other filters; the Suite dropdown keeps deliberate intersections. Incompatible suite/preset URLs explain the empty selection and link to the suite runs. - Show latest matching and latest full run links. Distinguish full, subset, unfinished, and unknown coverage from the recorded suite definition and execution counts. Keep failed full runs visible. - Retain the tested application URL and runtime mode in new history records. Older records show “Not recorded.” Browser-only journeys do not gain an inferred agent or model. - Add an index-only refresh command that can import a reviewed public campaign once. It rejects private Cloud bundles and does not replace retained campaigns or upload evidence. - Keep existing report URLs, test IDs, commands, and measurement timestamps stable. Campaign links carry suite/profile/environment; target/model/provider/runtime remain history-level filters. - Add unit and synthetic browser regressions, including mobile navigation, long source refs, sidebar filter reset, and empty-selection repair. Inset dropdown carets in the shared report styles. Document direct Node entry points without adding package.json shortcuts. - Add a prominent Reset all filters button, blue active-filter borders and tint, and visible active counts for the main and advanced filters. Reset clears all 13 history filters, sidebar search, URL state, and the attention view. Campaign-page filters use the same active styling. ## Verification - Post-rebase focused Vitest: 127 tests passed across 10 files. The latest follow-up selection passed 100 tests across 9 files. E2E support TypeScript passed after both changes. - The synthetic Chromium journey covers full and subset coverage, unfinished selections, correlated filters, browser-only QA selection, URL/back/reload, and desktop and touch navigation. The expanded journey verifies 13 active history fields, the collapsed advanced-filter count, reset of every field and sidebar search, and active border color changes that revert after reset. It makes zero provider calls. - Hub Python tests: 7 passed. The browser harness also exits promptly when Chromium is unavailable; its HTTP server closes on launch failure. - Real history verification covered 419 campaigns and 59 suites. All 418 previous campaign records remained deeply equal after importing the reviewed QA campaign. Publication used the authoritative history ETag. Only mutable index/history/pointer files were written; immutable campaign bundles stayed unchanged. - A live browser walkthrough passed from the two-family eval hub through QA smoke, Basic app journeys, target filtering, the retained report, screenshot gallery, history, and the logo link. The QA URL was clickable and the image loaded. No browser JavaScript errors occurred. The original QA normalized results and all three PNG hashes remained unchanged. - Live [E2E history](https://d1p6rlowie26tp.cloudfront.net/runner-e2e/?view=history), [core compatibility suite](https://d1p6rlowie26tp.cloudfront.net/runner-e2e/?view=history&suite=core-compatibility), and [QA smoke runs](https://d1p6rlowie26tp.cloudfront.net/runner-e2e/?view=history&preset=qa-smoke). - The live follow-up verified the incompatible Stock harness/QA-smoke URL, the repair link to 21 Stock harness campaigns, and sidebar navigation that clears preset/model filters. Dropdown carets were inspected in dark, light, and mobile views. No JavaScript errors occurred. - The final live follow-up verified reset of history controls, URL state, sidebar search, and the attention view, plus reload/back restoration and mobile touch reset. Active filter borders and tint were inspected in dark and light views. Campaign-page reset restored all three QA cases. All 419 history records and original QA measurement/evidence hashes stayed unchanged. - Full repository typecheck, test, build, and E2E CI gates passed on final head `1334cb79a5`. Both enforced aggregates, `ci / verify` and `ci / e2e`, completed successfully. Fresh Greptile review scored that exact head 5/5 with no unresolved review threads. ## Risks - Historical metadata can be missing. Unknown targets and runtime details stay explicit instead of being inferred from a worker environment. - Full suite coverage follows the saved definition and selection. Filtering to one model does not redefine that denominator. The latest full-run shortcut can cross coverage selections and says so in its detail text. - This changes shared report presentation. It adds no application runtime or database changes. The browser harness and index refresh do not invoke models. - Public history publication must keep its existing conditional-write and retained-evidence boundaries. Private bundles, raw logs, credentials, and screenshots are not part of this PR. ## Model Used - OpenAI Codex, GPT-6 family, with reasoning, code execution, and browser tools. The exact backend model ID and context-window size are not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Dotta
|
b25f3e2af8 |
fix(onboarding): show native proposal confirmation recipes (#18025)
## Thinking Path > - Paperclip lets people manage agent work through tasks. > - First-task onboarding asks a new agent to propose work before it starts. > - The user must accept a saved proposal before the agent can create its child task. > - A native agent can ask for approval in a question card, which does not record proposal acceptance. > - The skill names confirmation cards but does not show their native tool arguments. > - This change adds executable native proposal recipes and checks their real approval path. > - The task admission rules and Product E2E assertions stay in place. ## Linked Issues or Issue Description **What happened?** A fresh onboarding agent asked the user to approve a child task through an Approve/Revise question. The question looked like a decision, but its answer cannot authorize child creation. The first-task skill teaches native question calls for interviews and names proposal confirmation cards without showing the native mapping. **Expected behavior** The agent should create a confirmation card for a single-task proposal, or a checkbox card bound to the saved plan revision. Only the saved human acceptance should admit the child task. **Steps to reproduce** 1. Start first-task onboarding with a native agent. 2. Choose one task and ask the agent to propose a child before it starts work. 3. Have the agent use a question card with Approve/Revise options. 4. Answer Approve. Child creation remains blocked because no proposal confirmation was accepted. Related public context: #17940, #13517. ## What Changed - Show the native `request_human_input` arguments for confirmation and plan-bound checkbox proposals. - Explain that question answers supply scope and cannot grant proposal authority. - Resume the agent after either acceptance or rejection, and use new proposal keys for revisions. - Parse the shipped recipes and execute them through the real native tool authority and interaction service. - Verify that pending proposals, agent-only acceptance, and an answered Approve/Revise question cannot admit child work. - Use the existing isolated PostgreSQL test helper when a local test database is supplied. - Fix the required clean Grok npm-install check to prefix synthetic revision identifiers with `g`. The former all-numeric, leading-zero identifier was invalid SemVer and caused a reproduced `ETARGET` failure. Customer release versions are unchanged. ## Verification - Red: both new executable recipe checks fail against the preceding skill. - Green: 44 focused tests pass with zero skips against a task-owned PostgreSQL server. The actual tool authority and approval service are exercised. - Human acceptance of either recipe admits the child. Agent-only acceptance does not. Answering the question leaves the issue set unchanged. - Red/green real HTTP checks prove rejection saves the assignee continuation, starts no extra model run, and still denies child creation. - `git diff --check` passes. - Full repository `pnpm -r typecheck` and `pnpm build` passed on the final approval-guidance revision. - The first full local test attempt used an unmigrated external fixture and was stopped after the cause was confirmed. A new migrated fixture resolved that setup error; the second full attempt reached its explicit 10-minute limit before finishing. Both attempts are preserved and are not claimed as passing. The temporary owned PostgreSQL container was removed. The complete Linux CI workflow passed on the final PR head. - The exact synthetic-version failure was reproduced with real npm and an empty local packed registry. The new maintained regression keeps the invalid-version negative control and installs corrected numeric-hash and normal-hash controls. All 27 existing Grok helper tests and three offline registry-boundary checks pass. - Final head `52d0b92a00dcb0653b6fb48415cfaea979b8f21f`: [CI 40706933035, attempt 1](https://github.com/paperclipai/paperclip/actions/runs/40706933035) passed; 55 successful checks, two skipped and one neutral. Greptile is 5/5 for this exact head with no unresolved threads. The original failed build remains preserved. - Live fresh-workspace acceptance has not passed yet. It follows the reviewed testing rollout; this source test does not prove model compliance. ## Risks The change guides model tool selection. An agent can still ignore the skill. Product admission remains the authority, and the live journey must still verify human approval, native child execution, downloaded artifact bytes, context reuse, and cleanup. There is no schema or API change. The CI helper changes only temporary verifier versions and preserves the existing registry integrity, install, lifecycle, and isolation checks. ## Model Used OpenAI Codex `gpt-6.1-sol`, with reasoning, source inspection, tool use, and code execution. The host does not expose its configured context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge |
||
|
DottaandPaperclip
|
7d8bab773d |
perf: avoid full JIT and activity scans in Mine unread badges (#17973)
## Thinking Path > - Paperclip helps people manage AI agents and their work. > - Mine includes tasks the person created, was assigned, or participated in. > - Its unread badge ranks the visible personal top 100 before checking unread state. > - Correlated participation and privacy predicates can trigger expensive PostgreSQL JIT work. > - This change joins native participant IDs before ranking a narrow candidate projection. > - A partial actor index avoids scanning company activity history for each badge. > - Local representative scoped reads improve while dense unrestricted overhead remains explicit. ## Linked Issues or Issue Description **What existing behavior does this improve?** The Mine unread badge in the company sidebar. **Current behavior:** A fresh five-minute window recorded 402 ranking calls and 31.7 seconds of SQL time. Correlated predicates can spend more time compiling than executing. **Proposed behavior:** Build the same company-scoped participant set, use native issue keys, apply fresh authority before the top-100 limit, then retain the original canonical rank and second unread query. **Reason and benefit:** Reduce full JIT compilation and avoid an unindexed company-history activity scan. Realistic activity-only participation is included in the measurement. **Breaking changes:** No API or policy change. This PR adds one concurrent partial index migration. Related Mine rules in #16356 and #17756 are preserved. ## What Changed - Join distinct native participant IDs; guard exact legacy text before UUID conversion. - Skip redundant activity participation already covered by creation/assignment. Sort a narrow eligible candidate projection before canonical activity ranking. - Share existing pending-request and user-activity predicates with ordinary Mine reads. - Add partial `activity_log(company_id, actor_id, entity_id)` for user-authored issue activity. Register its idempotent concurrent migration with the existing top-level/invalid-index-repair runner. - Add real PostgreSQL coverage for visible top 100, unread/archive/source semantics, exact IDs, company scope and fresh grant/private-project access. - Add a pinned service benchmark with separate pre-index/indexed databases, identical data, full positional equality, ordinary-call timings, plans and real migration/cancellation proof. ## Verification - Fourteen focused PostgreSQL 17 tests pass (16.37 s), including assembled badge routes and grant revocation. Full workspace typecheck and build pass on the final production sources. - The earlier full local suite is incomplete: native PostgreSQL startup exhausted system semaphores, independently reproduced with empty initdb. All sixteen failed cases passed with task-owned Docker PostgreSQL 17. This is not counted as a full local-suite pass. Full Linux CI passes on final head `f44e46ce580e8345209d7165f57bd8bf696c517c`. Greptile and Paperclip Review are both 5/5 with zero unresolved comments. - Baseline `70c864633744cd96c0dc7394d631fc7119cd17d2`; source and DDL hashes retained. Two warmups and five alternating samples per arm/cell. Capture and EXPLAIN occur outside timed ordinary service calls. - All twelve 300k-history cells match complete positional rows in both SQL phases and final returned IDs. Half the personal tasks qualify through activity alone; the actor index contains matching and other-human rows. | Local service case | Before median | After median | Before / after | | --- | ---: | ---: | ---: | | dense / unrestricted / JIT on | 18.33 ms | 25.06 ms | 0.73x | | dense / scoped / JIT on | 909.30 ms | 221.10 ms | 4.11x | | sparse / unrestricted / JIT on | 13.64 ms | 5.84 ms | 2.34x | | sparse / scoped / JIT on | 719.07 ms | 10.79 ms | 66.64x | Small paths add at most 0.6 ms at the median. Dense unrestricted paths regress by 6.7–7.7 ms; dense scoped JIT-off is approximately neutral (218.34→223.24 ms). Sparse JIT-off paths improve too. This is not a uniform speedup. The actual baseline fails the compiler-work assertion. The final query avoids full JIT optimization/inlining and exceeds 1.5x where the old plan triggers those passes. Small/sparse controls meet the 25% or 2 ms allowance. Dense no-JIT cases are explicit diagnostics. The optional 60 ms absolute goal still fails at 221.10 ms on dense scoped reads: substantial privacy-predicate execution remains after compilation is removed. The failure is retained rather than counted as green. The actual normal migration runner upgraded 300k activity rows in 668.79 ms including inspection, index build/validation and journal work. Rerun took 21.83 ms; a deliberately cancelled build left an invalid index, and retry repaired it in 187.71 ms. Concurrent reads/writes completed; the final index was 188416 bytes. No migration was run on the hosted database. The million-row mixed-participation sparse control also passes all four visibility/JIT pairs with complete equality: unrestricted JIT-on 13.44→5.76 ms; scoped JIT-on 807.76→11.59 ms. The actor branch visits 36 indexed rows rather than scanning company history. Actual migration+runner took 255.06 ms; rerun 20.32 ms; cancellation repair/retry 241.73 ms. Concurrent reads and writes completed without failures. Reproduce: ```sh pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-mine-unread-badge.ts --baseline-ref=70c864633744cd96c0dc7394d631fc7119cd17d2 --history-decoys=240000 --verify-migration --assert-performance --evidence-dir=<directory> ``` Add `--fixture=sparse --history-decoys=940000` for the million-row control. The default harness starts disposable embedded PostgreSQL and ignores DATABASE_URL. Recorded evidence used PostgreSQL 17.11 through a scratch constructor provider because the host native runtime has semaphore limits. All normal database allocation/migration/client paths are retained. Source-only EXISTS/lateral alternatives were measured and rejected after company scans or sparse regressions. A time bound would lose historical participation, and existing indexes cannot select a user's activity. This justifies the exceptional partial index migration. The exact sample ranges, plans, failures, source hashes, migration footprint and feedback-loop timing are in `doc/plans/2026-10-27-mine-unread-badge-query.md`. ## Risks - Dense unrestricted reads are slightly slower, and dense scoped reads retain substantial authority work. The PR removes a compilation/lookup bottleneck rather than all badge latency. - The concurrent build scans activity history and can wait for long-lived transactions. Local build/cancellation timings do not guarantee the duration on larger or slower hosted databases. User-authored issue activity gains one narrow index to maintain. - Participation, visibility, ranking and legacy ID semantics must remain exact; the real database regressions exercise those boundaries. - Hosted improvement is unverified. The affected ranking SQL averaged 78.8 ms while HTTP badge requests averaged 752 ms without correlated traces. No merge or deployment is included. ## Model Used OpenAI Codex, GPT-6 family, with reasoning, tool use and local code execution. The runtime does not expose the exact model ID or context window to this agent. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
a91f78e7e0 |
perf(chat): reduce inbound queue discovery cost (#17971)
## Thinking Path > - Paperclip helps people manage companies of AI agents. > - External chat messages enter a durable delivery queue. > - Queue reads must preserve pending wake ownership and conversation order. > - Retained processed history increases the estimated cost of a correlated pending-action test. > - PostgreSQL 17 enables JIT for these small selectors and reads extra JSON in deadline discovery. > - This pull request uses a scoped pending-action set for open-work discovery. > - The benefit is less database work with the same selected deliveries and deadlines. ## Linked Issues or Issue Description **What existing behavior does this improve?** Inbound delivery deadline discovery and conversation-head selection. **Subsystem affected** Server chat queue discovery. **Current behavior** A correlated `EXISTS` inside an `OR` has a high cost estimate on retained completed history. PostgreSQL 17 already hashes its pending-action read, but the estimate still enables JIT. Deadline discovery also reads JSON from completed deliveries that have no pending wake. **Proposed behavior** Use `(company_id, endpoint_id, delivery_id)` membership in the pending-action set for the two open-work selectors. Keep ready-only evaluation on its original scoped `EXISTS`. Keep retry, lease, ordering, lifecycle and endpoint rules. **Reason and benefit** The local retained-history fixture reduces deadline selection time. The small control and the final head selection are neutral. Ready-only selection remains an unchanged control. **Breaking changes** None. No API, schema, migration, stored history, worker, timer or notification changes. Related work: #17647 and #15927 change inbound wake-notice scheduling and selection. This PR changes the separate conversation queue ownership predicate. ## What Changed - Use a fully scoped pending-action set for open-work discovery. Exclude null delivery IDs to preserve exact boolean results. Keep duplicate actions from multiplying deliveries. - Preserve the original ready-only query. - Add five real PostgreSQL tests and a reproducible baseline/treatment benchmark. Capture the complete real service SQL, parameters, positional rows and execution plans. - Document the query cause, rejected hypotheses, test results and limits in `doc/plans/2026-10-27-inbound-queue-query-proof.md`. ## Verification Pinned pre-fix baseline: `09ad9adce2cf4742944353a3707a2f908e9bb35e`. Local fixture: Linux ARM64 PostgreSQL 17.11, Node 26.4.0. It has 1500 completed deliveries and actions, two open deliveries, and 32 KiB of synthetic incompressible JSON per historical delivery. A separate control has 20 completed rows. The normal settings use JIT on with a 100000 cost threshold, 128 MiB shared buffers and 4 MiB work_mem. The final quiet retained-history pair measured these local SQL medians. Each arm has two warmups and five alternating samples. The small control is neutral. | Actual selector | Baseline | Treatment | Complete positional output | | --- | ---: | ---: | --- | | Next inbound deadline | 18.78 ms | 1.58 ms | Equal | | Conversation head | 5.71 ms | 6.18 ms | Equal | | Ordinary ready control | 9.23 ms | 8.33 ms | Equal ; unchanged | The deadline plan uses fewer shared buffers. Head physical reads remain unchanged. Deadline discovery avoids JIT on this fixture. The pristine head estimate is below the JIT threshold in both arms, and its timing is neutral. With JIT disabled, deadline discovery still improves; head and ready timings overlap. - Red: the unchanged baseline fails the explicit local fixture work budget for JIT and deadline buffer reads. Green: the treatment passes. There is no timing threshold in ordinary CI. - Five new PostgreSQL tests and six existing assembled chat tests pass. They cover scoped/null/duplicate pending actions, boolean results, payloads, head barriers, reactions, paused/attention/AgentMail exclusions, retry/processing deadlines, provider ordering, lease loss, lifecycle retry and edit/admission races. - Actual old/new ready selectors have equal error outcomes for unrelated malformed retry timestamps. An initial AND-only oracle comparison was inconclusive and is not defect evidence. - `pnpm -r typecheck` passes. - `pnpm build` passes. - Final head `f3cac298d8a8a5377127b0db7057c52019bbe426` passes the complete [CI workflow](https://github.com/paperclipai/paperclip/actions/runs/40625611887). Greptile is 5/5 on that head with no unresolved review threads. The earlier end-to-end cleanup lock-timeout failure passed on this revision. - The complete local suite was not run. The host's known native PostgreSQL startup failure exhausts SysV semaphores. Focused integration tests use a disposable Docker PostgreSQL constructor through the existing test seam. Complete final-head CI is required. Reproduce the real SQL comparison: ```sh pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-inbound-queue-discovery.ts --baseline-ref=09ad9adce2cf4742944353a3707a2f908e9bb35e --require-work-budget --retained-only ``` Use `--jit-off` for the diagnostic control. The default command uses the repository's bundled PostgreSQL version and records it in output. The reported 17.11 run used the existing test constructor seam with a disposable `postgres:17` container. Raw samples, complete queries, parameters and plans are emitted as JSON. The benchmark uses disposable synthetic data and does not read a live database URL. ## Risks - Query plans depend on data and PostgreSQL settings. These local measurements do not prove deployed speed or request latency. - Head selection still reads historical JSON. Action history remains stored and scanned. This is not a constant-time queue implementation. - The changed false branch has no retry timestamp cast. Ready-only parsing, claims and retries stay on their existing query. - Background SQL totals can overlap. Do not add these savings into an HTTP latency claim. The largest possible saving for an affected selection is bounded by its observed SQL time, about 49–53 ms on average in discovery. ## Model Used OpenAI Codex, GPT-6 family, with tool use and code execution. The runtime does not expose the exact model ID or context-window size. Three parallel agents own separate performance fixes; this PR contains only the inbound queue fix. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
f69dfc0e23 |
perf: use indexed latest-run probes for task interruption lists (#17969)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Task lists show the latest run's interruption state for each visible task. > - The old lookup applies a coalesced text expression to company run history. > - That lookup cannot use the separate binding indexes for a small task page. > - This pull request selects the newest run through those existing indexes. > - The benefit is less database work with the same interruption result. ## Linked Issues or Issue Description **What existing behavior does this improve?** The latest-run lookup used by task-list interruption enrichment. **Current behavior** The query scans company history for a coalesced native, first-class, or legacy task ID. A local two-task fixture inspects 1,506 runs to select six candidates. A recent hosted sample recorded about 110 ms per latest-run lookup. This is one part of the list handler, not its whole latency. **Proposed behavior** Probe each existing binding index for the newest candidate. Pick the newest candidate across the three binding branches. Read the same evidence fields in the same SQL statement. Preserve native precedence, first-class precedence, legacy text matching, company scope, and ordering ties. **Reason and benefit** The representative local 100-task page improves from 17.17 ms to 3.00 ms (5.72×). A modern 100-task page with 40 runs per task improves from 26.78 ms to 3.82 ms (7.02×). All positional SQL fields, row order, and full service results match. **Breaking changes** None. Stored history, API fields, visibility filtering, execution authority, and external-conversation state are unchanged. Related behavior: #16965, #17509, #17552. The duplicate search found no open PR for this query rewrite. ROADMAP.md has no duplicate planned feature. ## What Changed - Use three disjoint indexed candidate branches with a lateral newest-run selection. - Guard UUID parameters to preserve the old text matching for noncanonical input. - Keep final evidence projection in the same database snapshot. - Add real PostgreSQL work and contract regressions, a pinned-source benchmark, and a verification note. - Require the plan regression to find real run-table nodes. Record the deep-native-history follow-up without adding indexes or DDL. ## Verification - Baseline red: the real two-task lookup inspects 1,506 run rows and fails the work bound. The unchanged semantic oracle passes. - Treatment green: both PostgreSQL query regressions pass. They cover conflicting bindings, malformed JSON, empty values, noncanonical IDs, duplicates, foreign companies, missing runs, and timestamp ties across branches. - Seven paired PostgreSQL 17.11 cases pass complete positional SQL and full service equality. Each arm has two warmups and five alternating measured samples. Fixture payload sizes are random input bytes before Base64 encoding: 24 KiB becomes 32 KiB of stored text per noise field. - Representative 100 × 12 history: 17.17 ms (16.79–17.48) → 3.00 ms (2.64–4.28). Shared buffer hits: 30,033 → 3,400. - Modern 100 × 40 history: 26.78 ms (25.79–27.40) → 3.82 ms (3.47–4.78). - The benchmark command, all cases, plans, fixture limits, and feedback-loop timing are described in [the verification note](doc/plans/2026-10-27-task-interruption-query-performance.md). - Reproduce with `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-task-interruption.ts --baseline-ref=09ad9adce2cf4742944353a3707a2f908e9bb35e`. - `pnpm -r typecheck` and `pnpm build` passed at `20d46e3a16d10cf3c5399fc1211b17db6a4afdc6`. The current head has the identical production runtime; it adds one test assertion and documentation. - The focused query, real project-access repair/retry, and presentation suites pass: 19 tests across three files. The earlier plugin-sdk loading failure is resolved by normal build setup. - `pnpm test:run` was attempted, then stopped after repeated native embedded PostgreSQL bootstrap failures. The full local suite is incomplete. All 16 affected real-DB executor cases pass when rerun with disposable PostgreSQL 17 in Docker. - Linux CI passes the full server suites on this head. One unchanged Runner close fixture failed on the first CI attempt; its source tree is identical to the passing upstream #17439 baseline. One targeted job retry passed on the original head. Fresh checks run again for the test/documentation review fix. A separate macOS diagnostic fails the unchanged close variants too; that close issue is not diagnosed or counted as a local pass here. - The original head passed 55 checks with one neutral result and two expected Storybook skips. Greptile gave 5/5. The current head adds one nonempty-plan assertion and documents the deep-history follow-up. Its focused PostgreSQL suite passes 2/2. Final head `c1db106d7ab56b530af82d2bb0556feb0eda290b` is mergeable with all CI checks complete and passing, Greptile 5/5, and no unresolved review findings; the documented per-task history limitation remains an explicit follow-up. ## Risks - More index probes add about 0.43 ms to the one-run tiny control. The maximum 1,000-task page with one modern run per task is 9.99 → 14.11 ms, a 4.12 ms regression. Dense history pages save substantially more in the local fixture. - Each branch can still walk one task's own history. The first-class and legacy probes can discard native-bound rows after index lookup. The plan note specifies a 10,000-run follow-up, migration-free alternatives, and a measured index-build budget if indexes prove necessary. This PR does not make work independent of history depth. - Local SQL results do not prove deployed HTTP gains. The affected query averaged about 110 ms in the hosted sample; its critical-path contribution can be smaller because other enrichment reads run concurrently. - The main issue-list SQL and external-conversation gates remain separate work. This PR has no migration or startup job. ## Model Used OpenAI Codex, GPT-6 family, with code execution and repository tool use. The runtime does not expose the exact model ID or context window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
|
39279aa34f |
feat(ui): shared error, connectivity and query foundation for disconnects (#15754)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - The board reads API data and keeps the current task on screen. > - A restart or temporary network failure can interrupt those reads. > - Separate page error handlers can hide cached data or expose raw server codes. > - The app needs one connection state, readable error copy and a consistent read policy. > - This change provides that foundation and moves affected read surfaces onto it. > - The dependent write layer in #17960 preserves edits and submissions across the same failures. ## Linked Issues or Issue Description **What happened?** Temporary API failures could replace loaded content with errors. Different surfaces handled the same outage differently. Some showed raw server codes. A successful query to an unrelated service could also falsely report that Paperclip had recovered. **Expected behavior** Keep cached content visible. Show one connection banner during sustained trouble. Confirm recovery through the Paperclip health probe or its WebSocket. Retry transient reads and show readable errors for permanent failures. **Steps to reproduce** 1. Open a populated task or workspace page. 2. Interrupt API requests with a 503 or a dropped connection. 3. Keep an unrelated external request in flight. 4. Confirm that cached content remains and that the external result cannot clear the outage. 5. Restore the API and check that reads refresh. The write layer first merged into this feature branch in #15769 now has its own dependent PR, #17960. This PR contains the shared foundation and read changes. Merge this PR first. ## What Changed - Classify API errors and provide readable error copy, including non-JSON proxy responses and Retry-After handling. - Use one connection store and health probe with bounded backoff. Confirm recovery from the server health probe or live-updates WebSocket. - Pause reads while disconnected, retry transient errors, and refresh active reads after recovery. - Keep cached data visible through shared query views and affected page, panel and plugin read surfaces. - Add one app-level connection banner and support the pending-write count consumed by the dependent write layer. - Add a query-error rendering guard. Call it directly from the existing CI step. Update development documentation. ## Verification - Historical-chat lookup now follows the same ready/stale roster state as rendering. Three regressions fail before the fix; the six new cases pass after it, including working Retry, fresh data, no-cache error, permanent 403 and company isolation. All six focused query/recovery suites pass 200 tests. Final-commit CI passes and Apex is 5/5 with no unresolved findings. Canary Dry Run initially refused a preview version that was already on npm. It passed after one rerun with fresh registry metadata; no check was disabled. - Integrated current master and retained both transcript interest registration and connectivity reporting on socket open. The seven focused recovery/query suites pass 288 tests, and UI typecheck and production build pass. All 9,451 UI tests across 799 files pass on the merge commit. Final-commit CI passes and Apex is 5/5 with no unresolved findings. - Current head `70d9335d732219352ef0f98414439766bbcb33df`: all hosted CI checks pass. Apex is **5/5** on that head with no unresolved findings. The branch has no merge conflict. GitHub still requires code-owner approval. - The foundation passes 9,430 UI tests in 804 files. After integrating current master page paths, 190 focused page tests pass. - Two unrelated-success regressions fail before the latest fix. With the fix, 62 focused UI tests, all six guard tests and UI TypeScript pass. - The composed foundation and write layer pass **9,505 UI tests in 808 files**, UI TypeScript, a production UI build, and query-error, token and map guards. - A real Chromium journey on the composed source injects a 503 during title save. The loaded task stays visible; one reconnect banner appears; the edit saves after recovery; another task stays unchanged; the saved title survives reload. - Full recursive local checks were not repeated because Runner compilation exhausted local disk during another repair. Hosted checks must pass before merge. No new staging deployment is claimed. - Latest review follow-up: a cached CaseDetail now survives a failed HTTP 429 refresh. Its page regression fails before the fix and passes after it. Eighty focused foundation tests, UI TypeScript, build and guards pass after that change. - Earlier CI evidence is retained: several foundation jobs received a runner shutdown signal together. The write-layer head also failed an unchanged Runner descriptor close test with `provider_transport_failed`. The same Runner source passed that suite in the foundation run. New CI runs cover the updated heads; no check or assertion was weakened. ## Risks - Shared retry defaults change the time before transient read errors appear. Permanent errors still fail promptly. - ApiError.message now contains readable copy. Callers that compare raw codes must use ApiError.code. - Cache success can come from another service or local data. It must not confirm Paperclip recovery. - Durable writes are reviewed separately in #17960. This PR does not change database or mutation contracts. ## Model Used Anthropic Claude Opus 5.5 (`claude-opus-5-5`) produced the original work with tool use. OpenAI Codex, GPT-6 family, assisted with restoration, integration, tests and review fixes. The Codex session does not expose its exact serving model ID or context window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: devinfoley <139239+devinfoley@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Paperclip <noreply@paperclip.ing> Co-authored-by: Amy (Opus) <claude@users.noreply.github.com> |
||
|
DottaandPaperclip
|
2cd747d1b6 |
perf(server): batch connection install capability checks (#17945)
## Thinking Path > - Paperclip manages AI agents and their access to external connections. > - The connection grants screen describes which agents can receive an installation. > - Each candidate currently repeats authorization for every retained installation. > - A company installation makes this work grow with the square of the roster size. > - This pull request checks retained targets once inside the existing service owner. > - The benefit is fewer authorization calls with the same scoped capability results. ## Linked Issues or Issue Description **What happened?** The grants GET repeats connection preparation and retained-target checks for each candidate agent. A local 100-agent company-install fixture made 50,300 SQL calls and 10,000 target lookups in one capability-description operation. Open PR searches found no matching capability fan-out fix. **Expected behavior** The description must preserve retained-installation obligations and each candidate's permission. It should evaluate retained targets once per description. Writes must still authorize fresh. **Steps to reproduce** Run the committed benchmark with `--baseline-ref=f6bc653f7f745adaafa26393852fb2f5a99256a4`. It creates disposable synthetic PostgreSQL fixtures and loads the actual pre-fix service from Git. It compares every positional capability, reason, and operation. **Paperclip version or commit** Baseline `f6bc653f7f745adaafa26393852fb2f5a99256a4`. **Deployment mode** Built from source. Local real PostgreSQL proof. No hosted rollout. ## What Changed - Add `describeAgentInstallCapabilities` to connection management. Ordinary authorization and the batch share preparation, target checks, and fallback completion. - Use the batch only for direct saved non-AI connections and board actors. Complex resources, delegated actors, and unusual saved target IDs keep the ordinary evaluator. Fallback consumes the first preparation once. - Replace only the grants route's editable-agent loop. Keep configure/use checks, identity filtering, audience disclosure, and write authority on their existing paths. - Add scoped-grant, retained-target, error-order, freshness, fallback, full HTTP response, and real mutation regression coverage. Add a reproducible benchmark and evidence notes. ## Verification Real PostgreSQL 18.6 Linux Docker, Node 26.4.0. Two warmups and five measured samples per arm, in alternating order. These are service-component measurements with the same SQL-count instrumentation in both arms. They do not measure the whole hosted endpoint. | Fixture | Old median | Batch median | SQL calls old → batch | Target lookups old → batch | | --- | ---: | ---: | ---: | ---: | | 100 agents, company install | 24.753 s | 246.39 ms | 50,300 → 503 | 10,000 → 100 | | 100 agents, partial install | 6.339 s | 236.37 ms | 13,075 → 502 | 2,575 → 100 | | 1 agent, company install | 5.13 ms | 4.82 ms | 8 → 8 | 1 → 1 | All eight fixture shapes and every measured iteration had complete positional equality. One-agent timing ranges overlap. Full ranges, database settings, and limits are in `doc/plans/2026-10-27-connection-grants-batch-verification.md`. - **Passed:** `pnpm -r typecheck` and `pnpm build`, using pinned pnpm 9.15.4. - **Passed:** 116 focused tests: all 83 connection-management integration cases use real PostgreSQL and HTTP; 33 batch regressions execute real policy over synthetic query transport. The integration cases keep existing database creation, migrations, clients, services, and assembled HTTP server, with disposable Docker PostgreSQL. - **Passed:** complete grants GET parity across six human-role fixtures; real install PUT succeeds before revocation, then fresh GET and actual PUT deny after revocation without changing saved installs. - **Passed:** a real PostgreSQL red/green regression for empty saved target selectors, including candidate-specific decisions and later reached infrastructure faults. - **Passed:** independent old/new evaluator comparisons, including recursive and delegated fallback, run-context reuse, continuation, failure order, and principal/company isolation. - **Local full-suite limit:** `pnpm test:run` was attempted and stopped after native PostgreSQL fixtures could not initialize. An isolated failed native case reproduced that startup error. The host cannot allocate PostgreSQL SysV semaphores, including outside the sandbox. The focused suites use an existing constructor seam to start disposable Docker PostgreSQL. This is not a complete local full-suite pass. - **Passed:** all 16 native-runtime cases that failed to start PostgreSQL on the host, rerun with the disposable Docker provider. The full local workflow remains incomplete. - `git diff --check` passes. Greptile is 5/5 with no actionable findings on final head `41dcf3ae029b6c60999ca6208136399b0af663d0`. All final-head CI gates passed, including the full PR workflow. The two Storybook checks are expected skips because this change has no UI changes. Normal reproduction: ```sh node --import ./server/node_modules/tsx/dist/esm/index.mjs scripts/benchmark-connection-install-capabilities.ts --baseline-ref=f6bc653f7f745adaafa26393852fb2f5a99256a4 cd server pnpm exec vitest run src/__tests__/connection-install-capabilities.test.ts src/__tests__/connection-management.integration.test.ts ``` ## Risks Retained facts are observed once during one descriptive call. A mid-call revoke or deletion can leave a stale displayed choice. The next description and every mutation or human continuation read fresh authority. The batch returns no reusable authorization context. The bound applies to target evaluations in the eligible shape. Scoped grant payloads and hierarchy reads can still grow inside each policy decision. Complex resources and agent callers keep existing costs. Hosted latency improvement remains unverified. There is no migration, provider protocol change, new worker, persistent cache, or UI contract change. ## Model Used OpenAI Codex with code editing, execution, and test tools. Design, test, and independent review agents used exact model IDs `gpt-6.1-sol` and `gpt-6-astra` with xhigh reasoning. The host did not expose the primary session's exact model ID or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
09d4ca533e |
perf(server): return only sidebar failure badge candidates (#17935)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/95/changes/paperclip/95 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - The sidebar badge shows failures that need attention. > - The service reads the newest 200 runs before it selects one run per agent. > - Most of those rows cannot contribute to the failure count. > - This pull request selects failure candidates in SQL and returns four required fields. > - The benefit is less database response data with the same badge and access rules. ## Linked Issues or Issue Description Refs: #12770. Refs: #17756. Refs: #17683. These are related load, inbox, and badge reports. This implementation was written independently from the current service. It does not change which issue states count as failures. **What existing behavior does this improve?** The run read used by the company sidebar badge. **Current behavior** Every call transfers up to 200 rows with eight scalar fields. The application then selects the latest run per agent and applies status, Mine, dismissal, and access rules. **Proposed behavior** SQL selects the same latest run per agent from the same 200-row window. It then applies status and Mine rules. It returns only the run ID, scope, issue ID, and creation time. The service still checks dismissals and current issue access. **Reason and benefit** A mixed synthetic fixture reduces the response from 30,689 to 1,297 value bytes. The complete badge response stays equal. **Breaking changes** None. The endpoint response and the 200-run window stay the same. ## What Changed - Keep company and active-agent filters before the existing 200-row limit. - Rank each agent by the full creation timestamp in descending order, then by the observation ordinal. The timestamp order explicitly selects the newest run. Exact ties retain the first observed row. No ID tie-break is added. - Apply failure and Mine filters after ranking. A shared agent's success cannot expose an older failure. - Return four fields for dismissal and fresh issue authorization. Supply the already constrained company ID in the service. - Add real PostgreSQL regression fixtures and a service-generated SQL comparison harness. Document the command in `doc/DEVELOPING.md`. ## Verification - `pnpm -r typecheck` passed. `pnpm build` then passed in a separate sequential step. The normal Runner verifier checked 1,509 inputs and 1,555 outputs before reuse. - `PAPERCLIP_SIDEBAR_TEST_DATABASE_URL=<dedicated-test-url> pnpm --filter @paperclipai/server exec vitest run src/__tests__/sidebar-badges.test.ts` passed all 14 tests on PostgreSQL 17.11. - The 10 unchanged `inbox-dismissals.test.ts` route tests passed against the same real PostgreSQL server through a temporary local adapter for database startup and cleanup. The adapter was removed. The standard host run passed the 13 new tests and skipped those 10 embedded-PostgreSQL tests. - The first focused attempt could not import the unbuilt plugin SDK. The normal build supplied that package before the passing reruns. - `pnpm test:run` was attempted. It recorded 16 embedded-PostgreSQL startup failures in the unrelated native-session executor suite. Other database suites also reported startup failures and skipped. The owned command was then stopped with exit 143. This is incomplete local full-suite evidence, not a pass. Full remote CI passed on final head `df5b2c81af43c282d031cc37e6bc867054217df5`: 54 successful checks, 2 neutral checks, and 2 optional Storybook skips; no pending or failed checks. Fresh exact-head Greptile review is 5/5 with zero unresolved threads. - A fixture rollback test reproduced the review finding before the fix: a failed final approval insert left 2 companies, 44 agents, 2 issues, and 657 runs. After the fix, the same PostgreSQL constraint error leaves zero owned rows in all five tables. The test also confirms that more than 600 runs were written before the deliberate failure. - Fixture seeding now uses one transaction, including the submillisecond SQL insert. The 14 focused tests and all 10 benchmark cases pass after this repair. The server typecheck also passes. Production SQL is unchanged. - The final ordering repair passed all 14 real-PostgreSQL tests, all 10 pinned-baseline benchmark cases, and affected server typecheck and build, run sequentially. The plans now explicitly sort by agent ID, full creation time descending, then observation ordinal. Complete badge responses and candidate bindings remain equal. - One server-test shard was interrupted by its CI runner receiving a shutdown signal. Its test step was canceled without an assertion failure. A single targeted retry passed; the aggregate `ci / verify` gate also passed. No source change was needed for that interruption. - `git diff --check` passed. No schema, workflow, or lockfile changed. Reproduce the benchmark with current schema migrations and a dedicated synthetic database whose name ends in `_test`: ```sh PAPERCLIP_SIDEBAR_TEST_DATABASE_URL=<dedicated-test-url> PAPERCLIP_LOG_LEVEL=error SIDEBAR_REQUIRE_BYTE_REDUCTION=1 pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-sidebar-badges.ts --baseline-ref=bf3ef12f3a78d0b15a32ad57f7e16520fbf6cd5e ``` The harness loads the baseline service from that commit. Its authorization, privacy, and shared helper dependencies are unchanged. It captures the SQL each service executes. It compares complete badge responses and every returned candidate field. Fixtures cover more than 200 runs, shared agents, superseding successes, exact and submillisecond ties, Mine user variants, dismissal equality, private and deleted issues, missing issue bindings, terminated agents, and cross-company rows. Tests also cover absent actors, grants, revocation, deletion after a prior read, and index versus sort plans. Synthetic local SQL results from PostgreSQL 17.11 on Docker, Node 24.22.1, and postgres.js 3.4.9 follow. Each timing is the median of seven alternating warm samples after two warmups. The report includes all samples, ranges, settings, SQL, parameters, and `EXPLAIN (ANALYZE, BUFFERS, FORMAT JSON)` plans. | Fixture | Rows before → after | Value bytes before → after | Reduction | Median ms before → after | | --- | --- | --- | --- | --- | | Mixed, company-wide | 200 → 18 | 30,689 → 1,297 | 95.8% | 1.361 → 1.305 | | Mixed, user-1 | 200 → 9 | 30,689 → 676 | 97.8% | 1.785 → 1.391 | | Mixed, local-board | 200 → 6 | 30,689 → 414 | 98.7% | 1.463 → 1.232 | | All failed, 200 distinct latest agents | 200 → 200 | 30,298 → 13,778 | 54.5% | 1.277 → 1.237 | | All succeeded | 200 → 0 | 30,898 → 0 | 100% | 1.095 → 0.898 | | Empty | 0 → 0 | 0 → 0 | Neutral | 0.703 → 1.002 | Value bytes mean UTF-8 non-null cell values before Drizzle decoding. They exclude PostgreSQL protocol framing and are not wire bytes. The primary claim is less response data. Warm timing ranges overlap. Earlier all-failed samples were slower, and the empty control remains slower. SQL adds ranking work. No production latency improvement is claimed. The plans keep the limit below both window operations. ## Risks - SQL adds two bounded window operations. The all-failed control transfers fewer bytes but can take more SQL time. - Per-agent ranking explicitly orders by full creation time descending. Exact timestamp ties retain the observation ordinal. Real PostgreSQL tests compare complete candidates with the original helper under index and sort plans. - Mine rules now also have a SQL form. Tests cover undefined, null, empty, named, and local-board users, including empty responsible-user IDs. - Issue access stays fresh and fail closed. No authorization result is cached. No migration or stored run change is required. > For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and discuss it in `#dev` before opening the PR. Feature PRs that overlap with planned core work may need to be redirected — check the roadmap first. See `CONTRIBUTING.md`. ## Model Used OpenAI GPT-6 through Codex. The exact serving model ID and context window are not exposed in this session. The model used reasoning, code execution, SQL inspection, and GitHub tools. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
dc8cbdb819 |
perf(server): narrow native session resume database reads (#17934)
<!-- entire-trail-link-start --> https://entire.com/gh/paperclipai/-/trails/94/changes/paperclip/94 <!-- entire-trail-link-end --> ## Thinking Path > - Paperclip manages AI agents and their work. > - Native agents resume prior provider sessions from saved run checkpoints. > - Resume selection must respect every newer progress barrier. > - The prior query returned large context and profile JSON for up to 32 candidates. > - Most candidates need only process, status, binding, and checkpoint-presence checks. > - This PR selects the same source in one SQL snapshot and returns only its required payloads. > - This reduces database response bytes while preserving session authority and saved task context. ## Linked Issues or Issue Description **What existing behavior does this improve?** Native task/session resume selection in `findNativeSessionResumeRun`. **Current behavior** At baseline `bf3ef12f3a78d0b15a32ad57f7e16520fbf6cd5e`, a resume lookup reads complete context and profile JSON for up to 32 runs, then reads provider-event evidence. A synthetic 15-run history returns about 4.44 MB even though one checkpoint is selected. **Proposed behavior** A bounded CTE finds the first checkpoint or progress barrier. The same SQL statement returns the selected source's native input, checkpoint, tool fingerprint, wake envelope, canonical task brief, and process identity. All fallback barriers and the 32-run limit stay in place. **Reason and benefit** Rejected candidates no longer transfer saved prompts and execution inputs to Node. A single snapshot also prevents a gap between candidate selection and payload hydration. **Breaking changes** None to public APIs or stored history. This private service returns the fields consumed by runtime selection and checkpoint rebinding. It keeps complete required nested values and distinguishes missing brief fields from explicit JSON nulls. **Additional context** Checked `ROADMAP.md`. This is a targeted improvement to existing resume behavior. Public searches for native session resume projections found no duplicate. Related work includes #6573 (durable agent sessions) and #17011 (managed Claude resume opt-in); neither changes this database read. Provider-refusal recovery in #17676 is a separate concern and is unchanged. ## What Changed - Select the newest progress barrier within the existing exact company/agent/task/session and 32-run window in one SQL snapshot. - Return three required profile fields, the full wake envelope, and the saved task's ID, title, and description only for the selected source. - Preserve invalid non-null checkpoint barriers, provider evidence, process authority, terminal-status rules, timestamp precision, and exact string binding checks. - Add real PostgreSQL regression coverage and a reproducible baseline/treatment benchmark. Document the read and benchmark command. ## Verification - `pnpm -r typecheck` passed. `pnpm build` passed. The final server `tsc --noEmit` also passed. The later benchmark-only relocation passed `pnpm check:module-boundaries` and the real PostgreSQL proof; runtime and regression-test source are unchanged. - Related suites: `native-session-resume-projection.test.ts`, `native-session-resume.test.ts`, `heartbeat/retries.test.ts`, and `heartbeat-retry-scheduling.test.ts`: 78 passed and 54 skipped across 3 passed files and 1 skipped file. Embedded PostgreSQL/platform gates caused the skips. Resume orchestration and the 10 new SQL tests ran against isolated PostgreSQL 17.11. The final 10-test projection rerun passed. - New SQL cases cover all seven provider evidence kinds, each process authority field, all terminal statuses, wrong bindings and runtime, invalid checkpoints, null/scalar/array/object JSON, missing versus null brief fields, the 32-row bound, ordering ties, future barriers, submillisecond timestamps, and a change committed before the one selection statement. - `pnpm test:run` was attempted. The host's embedded PostgreSQL startup failed after five attempts. The broad run recorded failures/skips in embedded-database suites and was stopped after that concrete host limitation (exit 143). It is not a local full-suite pass. The full final-head CI matrix passes on `dcde4e3dac2173321d7be2406e7775af8411826a`: 55 successful checks, one neutral check, and two optional Storybook skips. Greptile completed at 5/5 on that exact head, with zero unresolved findings. The initial benchmark-location policy failure was fixed before this final run. - Reproduce the benchmark with `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-native-session-resume-reads.ts --baseline-ref=bf3ef12f3a78d0b15a32ad57f7e16520fbf6cd5e`. The optional `PAPERCLIP_SESSION_RESUME_BENCHMARK_DATABASE_URL` must name an isolated database ending in `_test`. - The harness imports the pinned baseline module, invokes both actual services, captures complete generated SQL and parameters, compares every retained consumer field, and captures `EXPLAIN (ANALYZE, BUFFERS, FORMAT JSON)`. Its incompressible wide fixture keeps a complete 96 KiB native input in the selected result. Raw PostgreSQL text DataRow frames are measured exactly; RowDescription, protocol command metadata, TLS, and TCP are excluded. | Synthetic local case | Baseline DataRow bytes | Treatment DataRow bytes | Reduction | | --- | ---: | ---: | ---: | | 15 runs, one selected checkpoint | 4,441,143 | 104,599 | 97.64% | | 15 unused bootstraps | 4,435,590 | 0 | 100% | | One small selected source | 6,395 | 6,319 | 1.19% | The committed-head benchmark uses two warmups and seven alternating samples per variant. Local round-trip medians and ranges were: | Synthetic local case | Baseline median (range), ms | Treatment median (range), ms | | --- | ---: | ---: | | Wide history | 20.32 (18.63–24.37) | 1.63 (1.40–2.21) | | Unused history | 18.52 (17.63–20.35) | 1.20 (1.13–1.56) | | Small selected source | 1.05 (0.95–1.70) | 0.78 (0.73–1.30) | The small case has overlapping ranges and a 1.19% byte change, so it is neutral. `EXPLAIN` server execution time was higher for the treatment (wide: 0.402 ms versus 0.028 ms across the two baseline statements). `EXPLAIN` does not serialize result rows; the new JSON presence checks do extra server work while avoiding large result transfer. These are local synthetic round-trip measurements, not production latency claims. The final proof ran on `dcde4e3dac2173321d7be2406e7775af8411826a` with no runtime source diff. ## Risks - Selection policy now has a SQL form. Regression tests preserve the pure selector's barrier semantics, including invalid non-null checkpoints and JavaScript Date precision. - Checkpoint-presence checks can still detoast candidate profiles inside PostgreSQL. The change reduces returned bytes; it does not eliminate all database JSON work. - The selected checkpoint, native input, and wake envelope can still be large because consumers require them in full. No new payload cap is introduced. - There is no migration, cache, background polling, ownership lock change, deployment, or merge in this PR. ## Model Used OpenAI GPT-6 through Codex. The host did not expose an exact runtime model ID or context-window size. The agent used reasoning, code execution, real PostgreSQL, and GitHub tools. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
3d07e06be7 |
perf(chat): project reply lineage run context (#17932)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - A run can continue a task that began in an external chat. > - The reply router follows durable lineage to find the authorized conversation. > - Each lineage hop reads the full run context, although it uses only nine keys. > - Wide prompts and workspace data make that read transfer far more data than the router needs. > - This pull request selects those keys inside PostgreSQL and preserves their exact JSON values. > - The benefit is less database transfer with the same reply destination and authorization checks. ## Linked Issues or Issue Description Refs: #17255, #17894. The first owns related chat review delivery. The second is an independent result-projection optimization. This PR has no dependency on another open PR. A public issue and PR search found no duplicate change. The roadmap does not list this query optimization as planned core work. **What existing behavior does this improve?** The server resolves the conversation that may receive a reply after one or more task continuations. **Current behavior** Each hop reads agent identity, responsible user identity, and the complete context snapshot. The snapshot can contain large prompts and workspace data that the router never reads. **Proposed behavior** Select only the nine context keys that this function consumes. Preserve missing keys, explicit JSON nulls, arrays, objects, and scalar values. Keep non-object contexts unchanged. Keep all durable authorization queries and the cycle-safe traversal. **Reason and benefit** A synthetic wide context needs 262,402 DataRow bytes before the change and 244 after it. A valid 24-continuation lineage needs 6,560,338 bytes before and 6,388 after it. The reduction occurs before PostgreSQL sends the result. **Breaking changes** None. The public API, stored context, reply bindings, tenant checks, and admission proofs retain their existing behavior. ## What Changed - Project an object context through `jsonb_each` and `jsonb_object_agg`. Retain only the nine consumed keys. Preserve all selected JSON values and key presence. - Keep the agent and responsible-user columns. Keep the company and run predicates. - Add 22 real PostgreSQL tests for routing, malformed JSON, strict tool-action receipt comparisons, tenant boundaries, 24 valid continuation hops, and an authorized cycle. - Add a synthetic benchmark that loads the pinned baseline service, captures its generated SQL, compares complete bindings and selected columns, and measures text-format PostgreSQL DataRow frames. - Run owned resource cleanup in nested `finally` blocks so fixture deletion or client close failures cannot skip cluster and scratch cleanup. Preserve the original benchmark failure alongside a cleanup failure. The complete consumer-key audit is: | Context keys | Consumer | | --- | --- | | `issueId`, `taskId` | Task identity fence, with the existing fallback order | | `wakeCommentIds`, `wakeCommentId`, `commentId` | Board reply and root comment provenance | | `source`, `sourceRunId`, `interactionId` | Durable continuation branches and tool-action receipt checks | | `endpointId` | Root destination fence | Missing `interactionId` stays absent. It does not become JSON null. This matters because the tool-action branch compares it strictly with the durable wake payload. The tests cover absent, null, string, number, and object values. They also remove the delivery receipt to prove that context alone cannot authorize a reply. ## Verification Baseline: `3cd3af5d75f200185ed4d9703a3145594a489b13`. Treatment: `3439ba7b714090d9f251816efec26622b3df223b`. PostgreSQL 17.11, Node 26.4.0. All data is synthetic. The benchmark holds imported helpers and fixtures identical and loads only the baseline `issues.ts` from history. It records generated SQL, parameters, plans, source SHA-256, and the initial worktree state before scratch creation. The committed final measurement reports a clean worktree and source SHA-256 `cd7e955179118fda12f6b2e981a17efff54137cbf241019785a3f8d16a7dcbb8`. Reproduce the benchmark: ```sh node --import ./server/node_modules/tsx/dist/loader.mjs scripts/benchmark-chat-reply-lineage.ts --baseline-ref=3cd3af5d75f200185ed4d9703a3145594a489b13 ``` The default creates disposable embedded PostgreSQL. `PAPERCLIP_CHAT_LINEAGE_BENCHMARK_DATABASE_URL` can name a dedicated external database that ends in `_test`. The harness applies migrations and removes its own fixture rows. It never uses an application database by default. Each wide run has 192 KiB of random bytes encoded as base64. This produces a 262,144-byte incompressible body. Two warmups precede seven timed samples. Baseline and treatment alternate order. Every scenario returns the same complete bindings. Every retained SQL column matches its baseline value, and the context matches the exact key projection. The acceptance check requires removal of every unrelated synthetic body byte. | Scenario | Baseline DataRow bytes | Changed DataRow bytes | Baseline SQL median ms (range) | Changed SQL median ms (range) | | --- | ---: | ---: | ---: | ---: | | small | 282 | 244 | 0.927 (0.900–1.455) | 1.027 (0.801–1.063) | | wide | 262,402 | 244 | 2.618 (2.440–4.831) | 1.490 (1.400–1.815) | | wide-24-continuations | 6,560,338 | 6,388 | 87.450 (80.103–130.197) | 43.103 (38.549–52.864) | These bytes are exact text-format DataRow frame sizes. They exclude row descriptions, command metadata, TLS, and TCP. Timings cover the captured lineage SQL, including all hops. They exclude the other authorization queries. The small control is neutral and timing ranges overlap on wider cases. An earlier loaded run had a small-control median of 1.15 ms before and 1.33 ms after. No deployed latency or billing improvement is claimed. The byte reduction is the primary result. Local checks: - `pnpm run preflight:workspace-links`: passed. - `pnpm exec vitest run server/src/services/chat-reply-lineage-context.test.ts`: 22 passed on the committed rebased head. - The existing `chat-interaction-continuation.test.ts` file: all 28 tests passed. These cover confirmation, verdict, connection, and invalid-authority paths. - Five selected cases in `chat-channels.integration.test.ts`: passed in a fresh dedicated database. They cover closed-choice question delivery and steering, Telegram confirmation, ambiguous close control, and both fresh-after-close authorization paths. - `pnpm -r typecheck` and `pnpm build`: passed again after the master rebase. - `node scripts/check-module-boundaries.mjs` and `git diff --check`: passed. - Cleanup repair at `3439ba7b714090d9f251816efec26622b3df223b`: all 22 focused PostgreSQL tests, the benchmark above, and server `tsc --noEmit` passed. Six controlled cleanup-failure cases confirmed later finalizers run; benchmark and cleanup errors remain visible together. The unchanged baseline failed the wide-result regression: 262,383 value bytes exceeded its 1 KiB bound. The treatment passes. Initial fixture setup needed a required conversation label and explicit foreign-key cleanup. The first full typecheck found a row-type mismatch in the new capture helper; that helper was corrected. The complete typecheck then passed. No production logic changed for those setup repairs. A prior assembled run reused the interaction test database. Its retained running fixtures caused a global milestone-count assertion to fail. Later reuse exposed retained bot registration and identity state. An attempted manual reset named a nonexistent table and did not complete. These failed attempts are not acceptance evidence. The final five-case assembled proof used a newly created and migrated database and passed every assertion. No existing assertion was changed. The full local `pnpm test:run` was not run. Focused tests used the coordinated disposable PostgreSQL server to avoid uncontrolled embedded-cluster concurrency on this host. The full final-head CI matrix passes on `3439ba7b714090d9f251816efec26622b3df223b`: 55 successful checks, one neutral check, and two optional Storybook skips. Greptile completed at 5/5 on that exact head, with zero unresolved findings. The existing interaction and assembled cases ran before the unrelated master rebase; the measured service source is unchanged. Focused tests, benchmark, full typecheck, and build passed on the rebased production-change head. The current cleanup-only head reran focused tests, benchmark, and server TypeScript validation; production service source is unchanged. ## Risks - Future lineage consumers must add any new context key to the SQL projection and the correctness oracle. - Selected values retain their complete JSON content. Non-object contexts pass through unchanged. Savings depend on unrelated object data being present. - PostgreSQL must still read the stored object to make the projection. The change reduces returned data, rather than removing all server-side JSON work. - Question delivery, tool action, connection, monitor, publication, company, and agent proofs retain their existing owners. No hop limit, cache, schema change, startup task, or deployment is added. - Benchmark usage and the result contract are documented here. No public command or API contract changes. ## Model Used OpenAI GPT-6 through Codex. The exact runtime model ID and context window are not exposed by this environment. Capabilities used: source analysis, code editing, shell tools, real PostgreSQL experiments, and automated tests. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
c3a9a8e92a |
perf(server): scope run identity reads to operation needs (#17931)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Tool and connection operations capture the current accepted requester. > - Capture locks the task and run, then reconciles pending steering. > - Most callers use identity or task fields from the returned run. > - The default query also transfers saved prompts, runner configuration, results, and logs. > - This pull request gives each production caller an explicit read scope. > - Narrow scopes keep their authority checks and return only selected columns. > - The Muse project path still reads its complete saved execution input. ## Linked Issues or Issue Description **What existing behavior does this improve?** Operation-time run identity capture for native tools, gateway sessions, project and workspace access, Slack, voice, connection management, and connection token admission. **Current behavior** These callers use the default full run read. Large saved execution documents cross the database connection even when the caller only needs the current accepted requester or task binding. **Proposed behavior** Use typed identity, task, broker, and project-execution scopes. Keep task-before-run locks, company and agent predicates, active status validation, and pending-steering reconciliation. Return the complete saved native input only where project admission uses it. **Reason and benefit** A synthetic PostgreSQL 17.11 fixture matches a wide run with a 142 KB context and a 77 KB runner profile. Its locked run DataRow falls from 225,101 bytes to 190–605 bytes for authority reads. The project scope returns 66,004 bytes because it must retain a complete 64 KB execution input. These are local DataRow measurements, not a forecast of production billing or latency. **Breaking changes** None to public APIs, saved data, or authority policy. Internal narrow return types omit columns that were not selected. The default full compatibility read and the existing GitHub-credential and authentication scopes keep their current contracts. **Additional context** The public issue and PR duplicate searches found no matching open change. This is one independent query optimization. It does not require other performance PRs. ## What Changed - Add typed scopes for accepted identity metadata, task authorization, token attribution, and saved project execution input. - Assign explicit scopes at all production `captureRunIdentity` call sites. - Keep the full native execution input for Muse project admission while omitting unrelated profile, context, result, and log data. - Narrow the broker's initial read and the existing Slack and connection-management historical reads to the same required fields. - Extend steering and acknowledgement regressions to all new scopes. Add real PostgreSQL projection, malformed JSON, tenant and agent, session-generation, and saved-input coverage. - Add a service-generated SQL benchmark with exact raw DataRow accounting, retained-field equality, plans, and alternating transaction timings. - Document the scope contracts and reproducible benchmark. ## Verification - Baseline: `623f1dc289a14a699ebbe3f09b94179d38fc2ae2`. - `vitest run server/src/__tests__/run-identity-projections.test.ts server/src/__tests__/run-identity.test.ts`: 54 tests passed on an owned disposable PostgreSQL 17.11 database. - The tests retain task-before-run and audit foreign-key lock behavior. Every scope rejects uncertain steering, accepts its durable acknowledgement, and preserves requester replay ordering. - A full-row mutation makes the narrow identity regression fail. The restored treatment passes. - `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-run-identity-projections.ts --baseline-ref=623f1dc289a14a699ebbe3f09b94179d38fc2ae2` captures the real locked SQL for both versions. - Every retained column and selected JSON field matches the baseline. Identity contexts match. The complete stored run remains unchanged. Capture keeps the same five queries. | Scope | Before DataRow bytes | After DataRow bytes | Reduction | | --- | ---: | ---: | ---: | | Identity | 225,101 | 190 | 99.92% | | Task authorization | 225,101 | 410 | 99.82% | | Token broker | 225,101 | 605 | 99.73% | | Project execution input | 225,101 | 66,004 | 70.68% | - DataRow bytes include row framing and raw text-format column values. They exclude response metadata, TLS, and TCP. - Two warmups and seven alternating samples measure the complete capture transaction. Several timing ranges overlap. The primary claim is byte reduction; production latency is not verified. - Initial fixture runs exposed an empty external schema and an incomplete synthetic conversation binding. The test setup now applies migrations and supplies the required user/state fields. The final 54 tests passed. - `pnpm -r typecheck` and `pnpm build` passed locally. - Four existing skill-setup tests passed. Four assembled connection-management/native-tool consumer cases passed with real JWT/API-key authentication. These include per-request identity freshness and task-scoped activity authority. - A CI-only stale mock assertion expected two capture arguments; it now verifies the explicit identity scope. A repository-wide capture mock audit found no other stale argument assertions. - Review follow-up: the projection suite uses the existing embedded-support guard while always enabling an explicitly supplied external database. The audit-lock regression merges URL query parameters and retains existing PostgreSQL options. - `vitest run server/src/__tests__/run-identity-projections.test.ts server/src/__tests__/run-identity.test.ts server/src/__tests__/company-skills-routes.test.ts server/src/__tests__/project-skill-setup-identity.test.ts`: 123 tests passed. The external PostgreSQL URL included both `application_name` and `options`; lock tests verified `lock_timeout=1s` and retained `statement_timeout=5s`. - Full local tests are not run because the host's embedded PostgreSQL setup is unavailable. Greptile reviewed `06e1dd53a3fb24df74d5fbfb2ad397f291f1e30f` at 5/5 with zero unresolved threads. The full final-head CI matrix passes. All 58 checks are terminal: 55 success, one neutral, and two optional Storybook skips. - No migration, lockfile, workflow, new polling, or cache change. ## Risks A narrow read must retain each caller's actual fields. The caller audit includes Slack's wake-request fallback, conversation generation, legacy broker attribution, and the Muse parser's complete native input. JSON selection retains value types, including malformed legacy values. The saved-input scope preserves missing-key behavior. The default full read still returns the complete stored row for compatibility. Locks and live authority checks are unchanged. ## Model Used OpenAI GPT-6, accessed through Codex. The exact runtime model ID and context-window size were not exposed in this session. Used repository inspection, code execution, and PostgreSQL validation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
56bc470ab2 |
perf(server): project continuation history result fields (#17930)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - A continuation carries completed tool receipts so the agent can avoid repeat actions. > - It also checks saved conversation and workspace recovery fields before choosing a prior terminal run. > - The history query transfers every result field for every matching run. > - Native output and unused receipt request fields can make this read large. > - This pull request selects the fields those consumers use and keeps each completed receipt result in full. > - The benefit is less database transfer with the same history, recovery decisions, and completed-action evidence. ## Linked Issues or Issue Description **What existing behavior does this improve?** The history read in `buildExecutionContinuation`. **Current behavior** The query returns complete `result_json` values for all matching runs. Consumers read completed receipt state, operation ID and result, plus conversation and workspace recovery controls. Large native output and receipt request fields cross the connection without being used. **Proposed behavior** Select only those result fields in PostgreSQL. Keep typed policy and recovery values. Keep all eligible completed receipts and their complete result values. Retain missing versus null receipt results, run and receipt order, company/task/agent scope, and the source-run exception. Keep the full retained history without a new limit. **Reason and benefit** The synthetic wide-output fixture removes 99.879% of this query's DataRow bytes. A required 2 MiB receipt-result control saves only 0.0074%, because that complete result is needed. Actual savings depend on result contents; the wide-output result is not a forecast for all deployments. **Breaking changes** None intended. No schema, API, migration, workflow, history limit, or receipt truncation changes. **Additional context** Related: #17481 projects a separate recovery read, and #17896 narrows explicit continuation admission. This PR covers the continuation history query and has no dependency on the other byte reduction PRs. ## What Changed - Project completed tool receipts to their state, operation ID and complete result. Omit fields that the continuation does not read. - Project typed conversation/workspace controls and the recovery schema. Keep the shared policy helper unchanged. - Add real PostgreSQL comparisons for complete continuation envelopes, ordering, scope, legacy JSON, recovery flags, and receipt result preservation. - Add external isolated database support to the existing continuation tests and document a reproducible query benchmark. ## Verification - Pinned baseline: `3cd3af5d75f200185ed4d9703a3145594a489b13`. - PostgreSQL 17.11 with synthetic local data. The harness captures SQL from the real baseline and treatment services. Imported helpers remain unchanged. - Command: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-continuation-history.ts --baseline-ref=3cd3af5d75f200185ed4d9703a3145594a489b13`. Set `PAPERCLIP_CONTINUATION_HISTORY_BENCHMARK_DATABASE_URL` to a disposable database ending in `_test` to use an external server. The default uses disposable embedded PostgreSQL. - The baseline was measured before the production edit with `--baseline-only`. The treatment compares complete envelopes, every retained scalar column, completed receipts, and shared policy decisions. The wide unused-output fixture must remove more than 90% of DataRow bytes or the benchmark fails. - Wide unused-output fixture: 25 runs with independent incompressible 192 KiB native output and 32 KiB unused receipt request fields. DataRow bytes: 5,744,950 to 6,925 (99.879% less). - Required-result control: one complete 2 MiB receipt result. DataRow bytes: 2,097,556 to 2,097,401 (0.0074% less). The whole result is checked for equality. - Exact byte boundary: PostgreSQL text-format DataRow frames, including field length prefixes. Excludes RowDescription, command metadata, TLS, TCP, and other service queries. - Seven measured query round trips per variant follow two warmups and alternate order. Final-head wide-output median: 32.86 to 4.41 ms (ranges 26.24–35.54 and 3.62–5.21 ms). Required-result median: 17.01 to 20.60 ms (ranges 16.30–32.89 and 18.85–22.48 ms). The required-result control shows projection overhead and overlapping timing ranges; it has no latency improvement claim. These are local measurements, not deployed latency results. - `PAPERCLIP_EXECUTION_CONTINUATION_TEST_DATABASE_URL=<isolated_test_database> pnpm exec vitest run server/src/services/execution-continuation.test.ts server/src/services/execution-continuation-history-read.test.ts --maxWorkers=1`: 55 tests pass in PostgreSQL 17. - Cases cover numeric receipt key order, empty operation IDs, malformed receipt entries, missing versus null results, complete large results, malformed root JSON, all closed workspace restore failure codes, invalid recovery schemas, cancelled admission rows, and company/task/agent/source-run boundaries. - `pnpm -r typecheck` and `pnpm build` pass on final head `d665d9f5c47b237c81b9cc80fdec1d39ad90754f`. The final-head focused PostgreSQL run passes all 55 tests. The final benchmark records that SHA, an empty source diff, and passing equivalence/byte assertions. The first typecheck found optional array accesses in new test assertions; explicit required-array assertions fixed them. Local full `pnpm test:run` is not claimed because the host embedded PostgreSQL runtime is unavailable; the full test matrix runs in CI. Exact-head Greptile review completed success at 5/5 with no actionable findings or unresolved threads. The full final-head CI matrix passes. All 58 current-head checks are terminal: 54 success, two neutral, and two optional Storybook skips. One original server shard ended with a canceled test step and no assertion failure; its single targeted retry passed. No product change was needed for that interruption. ## Risks - Future continuation consumers must add any newly required result field to this projection. Existing shared policy semantics stay in the helper. - Completed receipt results can still be large and can dominate transfer. This PR preserves them without a cap. The receipt-heavy control adds JSON projection work for only 155 bytes of savings, so it can be slower despite equal output. - JSON malformed-value and property-presence rules affect action evidence and recovery. Real PostgreSQL comparisons check those decisions and complete output. - Production transfer and latency improvements need a fresh deployed capture. Nothing was merged or deployed by this work. ## Model Used OpenAI GPT-6 through Codex. The exact runtime model ID and context window are not exposed by this session. The work used repository tools, code execution, and real PostgreSQL validation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
f60ac29cf1 |
fix(runner): migrate Grok Build with native reasoning effort (#17922)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Harness selection maps supported agents to Paperclip Runner. > - The native Runner already supports Grok Build, but its migration mapping excludes Grok. > - Grok agents also have a thinking-effort setting that the native execution contract does not carry. > - This pull request adds the migration path and preserves explicit Grok effort through startup and resume. > - Operators can upgrade an existing Grok agent without silently losing its model, account, or effort. ## Linked Issues or Issue Description **What happened?** Existing Grok Build agents have no Runner selector or upgrade action. Adding only the mapping would route conversion through the Codex fallback and lose Grok execution semantics. Native execution also ignores the legacy reasoning-effort field. **Expected behavior** A qualified Grok agent can explicitly upgrade. The saved model, account, instructions, history, and supported effort remain intact. Unsupported legacy settings fail before saving. The provider must acknowledge the selected effort before inference. **Steps to reproduce** 1. Open the runtime page of an existing Grok Build agent on a qualified Linux x64 or macOS ARM64 computer. 2. Observe that the Runner selector and upgrade action are absent. 3. Compare the native execution input with the saved legacy reasoning-effort setting. Related work: #13882 introduced native Grok Build. #15422 covers the earlier default-runner work. This change extends the current shared selection contract. ## What Changed - Add the Grok harness mapping, qualified platform limits, upgrade action, and company-scoped computer qualification. - Translate Grok settings to ACPX without changing the selected account or explicit model. Resolve the legacy default model to the native default. Preserve the effective legacy search policy. Reject incompatible command, argument, permission, and turn-limit settings. - Add native execution input v10 for explicit Grok effort and search policy. Preserve historical schemas. Enforce disabled search through Grok requirements and bind it across startup and recovery. - Apply and verify ACP `reasoning_effort` before prompting. Bind the value to provider descriptors, checkpoints, runtime identity, and resume checks. - Use the Grok catalog and effort choices in the task composer. Apply task model and effort overrides while retaining the agent account and search policy. Update runtime controls, portability/default handling, documentation, and capability maps. - Advance Pi and Copilot source attestations for changed shared transport and recovery code. Keep historical identities intact and update all Rust integration fixtures. ## Verification - Migration and review regressions were reproduced before the fixes. Coverage includes the Grok mapping, default model, explicit effort, implicit search policy, task overrides, and incompatible legacy settings. The broad migration, route, immutable-input, and composer check passed 478 tests. The final catalog and task-input check passed 135 tests. - Runner startup, recovery, search policy, and profile checks passed 361 tests, with one platform skip. The final source-identity, search-admission, and transport event-pump check passed 89 tests. - All 757 Rust workspace tests passed for the broad implementation. After the final source attestation update, all 55 ACPX profile/session tests passed. - `pnpm -r typecheck` and `pnpm build` passed for the implementation. Server and Runner typecheck passed after the final production change. Server `tsc --noEmit` passed after the final test-fixture change. Docker context integrity, token gates, maps, generated protocol data, and replay goldens passed. - The v10 startup fixture initially used an ambient local Grok login and failed on CI. An explicitly empty credential home reproduced all four failures. The fixture now supplies a synthetic key and a private empty home. A separate case checks that missing credentials fail closed. All seven selected Grok v10 and Claude v9 startup cases pass without host credentials. - The full local suite could not complete because embedded PostgreSQL failed before database-backed fixture setup. The broad executor-suite attempt hit the same limit and was stopped. Neither attempt is counted as passing. Current-head CI must verify those paths before merge. - Earlier CI found missing Rust input fields, stale source attestations and exact fixture pins, a missing Docker context entry, and an outdated import expectation. Those failures were fixed and checked locally. The CI logs retain the original failures. - Different earlier CI runs also failed unchanged cleanup pagination, device-login timeout, and chat lost-response recovery cases. Later attempts passed the cleanup and login cases without changes to those implementations. Their causes remain unproven. The original evidence remains in [the first run](https://github.com/paperclipai/paperclip/actions/runs/40563402840), [the login run](https://github.com/paperclipai/paperclip/actions/runs/40571128328), and [the chat run](https://github.com/paperclipai/paperclip/actions/runs/40573072163). A passing retry does not establish a root-cause fix. - Both reviews rate current head `16fba9b4cdf12bc7945e6628912e79da42959f35` 5/5. All current-head CI gates pass in [run 40576580699](https://github.com/paperclipai/paperclip/actions/runs/40576580699). All 60 check runs completed cleanly, including the four intentional skips and the neutral 5/5 Paperclip review. The legacy security status is successful. No review threads remain unresolved. - Staging migration and a real shell task are pending deployment. This PR does not claim full Grok provider qualification. Reviewers can verify the operator path by upgrading a qualified Grok agent, checking its saved model and effort, and running a shell task through native execution. ## Risks - Explicit effort and search policy require the matching Runner and sidecar build. Older immutable execution inputs keep their existing schemas. - Providers that omit or reject the selected effort fail before a prompt. Unsupported effort does not fall back silently. - Existing agents retain their saved runner until an explicit change. New Grok agents use native execution only on qualified platforms. - Grok token and cost reporting remain unverified, as before. ## Model Used OpenAI Codex, GPT-6 family. The exact serving model ID and context-window capacity are not exposed in this session. Used reasoning, source editing, shell execution, and browser inspection. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes: #` / `Refs: #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and the focused suites pass; the full local suite has the environment limitation recorded above - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
f6ba9afab0 |
perf(server): narrow explicit continuation run reads (#17896)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Recovery must prove that old execution stopped before it admits new user input. > - Explicit continuation reads the source run before and after it takes the run lock. > - Both reads transfer the saved prompt and the full Runner profile. > - The preflight needs only identity, comment, adapter, and workspace receipt fields. > - This pull request narrows those reads and keeps the complete locked profile for session proofs. > - The benefit is less database transfer with the same admission and recovery rules. ## Linked Issues or Issue Description **What existing behavior does this improve?** Explicit native continuation admission after a failed or stopped run. **Current behavior** The preflight and locked reread select complete heartbeat run rows. Large saved contexts and execution profiles cross the database connection twice. **Proposed behavior** Read only required scalar fields and typed context keys. Read only historical adapter identity and workspace sync receipts before the lock. Keep the complete Runner profile in the locked reread. Keep lock order, company scope, text lookup of historical run IDs, stop proofs, queued input, and replay rules. **Reason and benefit** A synthetic PostgreSQL 17 fixture removes 74.9% of the two queries' DataRow bytes. This result is local query evidence. It is not a claim about deployed latency or monthly transfer. **Breaking changes** None intended. No schema, API, migration, workflow, or dispatch contract changes. **Additional context** Related: #17491 narrows the separate execution blocker read path. This PR covers explicit continuation admission. The PR has no dependency on the other byte reduction PRs. ## What Changed - Select required continuation fields and typed context keys instead of complete run rows. - Read a small historical adapter/workspace profile during preflight. Preserve the full locked profile for exact suspended-session and quarantined-cleanup proofs. - Narrow four helper input types to their actual field requirements. Their runtime behavior is unchanged. - Add real PostgreSQL admission comparison coverage and a reproducible baseline/treatment benchmark. - Document the query boundary and benchmark command. ## Verification - Baseline: `663fe667d66d3863d0919a8724f1d99500feef0b`. - Disposable PostgreSQL 17.11. One source run has a 192 KiB incompressible context and two 96 KiB profile payloads. - Command: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-continuation-run-reads.ts --baseline-ref=663fe667d66d3863d0919a8724f1d99500feef0b`. Set `PAPERCLIP_CONTINUATION_BENCHMARK_DATABASE_URL` to a disposable database ending in `_test` to use an external server. - The harness captures SQL generated by both real service versions. Baseline imports use current helpers; only their TypeScript input types changed in this PR. Their runtime behavior stays fixed. - Preflight DataRow bytes: 394,067 to 466 (99.88% less). Locked reread: 394,067 to 197,168 (49.97% less). Combined: 788,134 to 197,634 (74.92% less). - Byte counts cover exact PostgreSQL text-format DataRow frames. They exclude RowDescription, command metadata, TLS, TCP, and other queries in the service. - Seven measured samples per query follow two warmups. The order alternates. Final-head median preflight time: 6.53 to 2.28 ms (ranges 5.34–7.41 and 1.84–4.46 ms). Median locked read: 8.06 to 5.26 ms (ranges 5.94–13.84 and 4.61–6.81 ms). These are warm local SQL measurements under shared machine load. The deterministic byte counts are the acceptance criterion; these noisy timings do not establish a deployed latency gain. - The harness compares the admitted continuation, blocked reasons, every retained scalar column, kept JSON, and the full locked profile. It fails unless the fixture removes more than half of the two reads' DataRow bytes. It records SQL, parameters, revision identity, samples, and EXPLAIN plans. - All 291 continuation tests pass together on the final head with no skips in a clean isolated PostgreSQL 17 database. This includes all seven new projection comparisons and six malformed/missing JSON cases. - Edited server TypeScript check passes: `node node_modules/typescript/bin/tsc --noEmit -p server/tsconfig.json`. - `pnpm -r typecheck` and `pnpm build` pass on the final head. The normal server build verifies the complete pinned Copilot closure. Local full `pnpm test:run` is not claimed: host embedded PostgreSQL startup limits prevent that gate. The isolated PostgreSQL regressions pass; the full exact-head test matrix runs in CI. The targeted exact-head test retry passes all 4,086 tests in 98 files. All CI verification checks and the dependent aggregate are complete and successful on head `47c1b5f7d17a58310f7a3ebd494fdcaa814777c2`. Greptile completed cleanly on head `47c1b5f7d17a58310f7a3ebd494fdcaa814777c2`: 5/5, no findings, and no unresolved review threads. - Local setup evidence: embedded initdb could not load its ICU data library. Tests therefore use a dedicated PostgreSQL 17 container. A run against reused test fixtures produced stale sweep failures; a new clean database passed all 285 cases. The first comparison mock replaced an unrelated aliased query; it was narrowed to the two admission reads before the clean pass. - The first [server test shard 6/12](https://github.com/paperclipai/paperclip/actions/runs/40529210562/job/122277218565) failed one existing workspace-recovery case: a third export stayed `scheduled` at its 20-second poll deadline; 4,085 other tests passed. The exact case passes on the pinned baseline and final head with the same PostgreSQL 17 fixture. Temporary spies recorded zero calls to both changed admission functions in either replay. Other touched helpers have type-only changes. This supports one timing-flake retry; the exact scheduler delay remains unproven. The single [targeted shard retry](https://github.com/paperclipai/paperclip/actions/runs/40529210562/job/122285744135) passes all 4,086 tests in 98 files. The exact case completes in 17.25 seconds. The dependent aggregate is complete and successful. No timeout or unrelated runtime code was changed. - The separate Paperclip Review check ended `NEUTRAL` without a final report. The completed source review is the exact-head Greptile 5/5 result above. ## Risks - A future admission check may need an additional saved context key. Keep the projection and helper contracts in sync with that consumer. - Exact suspended-session and quarantined-cleanup proofs still require the complete Runner profile. This PR retains it under the lock rather than weakening those checks. - JSON values retain their types and missing keys stay absent. Malformed-context comparisons preserve the existing controller hold. - These measurements use synthetic local data. Deployed transfer savings require a fresh capture after rollout. Nothing was deployed. ## Model Used OpenAI GPT-6 through Codex. The exact runtime model ID and context window are not exposed by this session. The work used repository tools, code execution, and real PostgreSQL validation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
6e5b9794d6 |
perf(github): avoid duplicate webhook body reads (#17894)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - GitHub webhooks enter a durable action queue before the worker processes them. > - A worker must check retry eligibility and win an atomic claim before it replays a delivery. > - The current eligibility read returns the complete webhook body, and the successful claim returns it again. > - This pull request reads small eligibility fields first and returns the body only to the successful claimant. > - The worker keeps its retry, ownership, runtime, credential, and recovery checks. > - The benefit is less database transfer without removing durable replay data. ## Linked Issues or Issue Description Refs: #17666, #15981. Those changes own the existing recovery scheduler and ingress module. This PR changes the result projections and validation order within that module. It has no dependency on another open PR. A public issue and PR search found no duplicate change. **What existing behavior does this improve?** The durable GitHub webhook worker reads action eligibility and replays claimed deliveries. **Current behavior** The first read returns all action columns and the complete webhook body. A successful claim returns them again. An active claim or future retry also downloads a body that this worker cannot process. **Proposed behavior** Read only ID, status, result, and update time for eligibility. Return ID, endpoint ID, and payload from the successful atomic claim. Validate that claimed payload before replay. **Reason and benefit** A synthetic 131,105-byte webhook body produces about 263 KB of response values before this change. The changed service returns about 132 KB for successful processing and fewer than 150 bytes for skipped work. This removes duplicate body transfer. **Breaking changes** There is no public API or schema change. Malformed active or delayed rows now remain under their current claim or retry deadline. The eligible claimant validates and fails them later. A worker that does not own the claim can no longer clear their payload. ## What Changed - Use a small projection for initial action eligibility. - Return the replay body only from a successful claim. Keep the existing status, stale-time, and previous-attempt guards. - Validate the row returned by the claim. Fence invalid-payload settlement by kind, processing status, and the claimed attempt. - Add 19 real PostgreSQL regressions for complete replay bodies, skipped work, retries, stale-claim recovery, contention, corrupt bodies, claim loss, replacement bodies, and payload retention. - Wait for the saved chat authentication link in one existing async UI test. Keep its exact destination and all other assertions. This repairs a second independent CI test failure without changing product UI. - Accept the two precise invalid-repository diagnostics from Git 2.55 and 2.56 in one existing test. This repairs an independent CI compatibility failure. It changes no Git expiration service behavior. - Add a reproducible benchmark that loads the real baseline service from an explicit commit. It captures generated SQL and parameters, query plans, response-value estimates, timings, and complete valid-work behavior comparisons. ## Verification - Baseline: `663fe667d66d3863d0919a8724f1d99500feef0b`. - PostgreSQL 17.11, Node 24.22.1. The fixture uses synthetic random body content. No hosted database or provider was changed. - `pnpm exec vitest run server/src/services/chat-channels/github/webhook-ingress.test.ts`: 19 passed with a dedicated disposable PostgreSQL database. - Existing assembled GitHub tests: recovery from pre-ingress failure, scheduler retry and restart, exact terminal redelivery, runtime pause/resume, recovery tombstones, and credential rotation passed. - `node scripts/check-module-boundaries.mjs` and `git diff --check`: passed. - Benchmark: `node --import ./server/node_modules/tsx/dist/loader.mjs scripts/benchmark-github-webhook-ingress.ts --baseline-ref=663fe667d66d3863d0919a8724f1d99500feef0b`. The default uses disposable embedded PostgreSQL. `PAPERCLIP_TEST_DATABASE_URL` can select a dedicated migrated test database with a name that ends in `_test`. - Ten timed samples after two warmups, with alternating baseline/treatment order. The main acceptance criterion is response size, rather than a timing threshold. The 0.51 ratio bound is deterministic for this fixed synthetic fixture; it is not a promise for arbitrary webhook bodies or result metadata sizes. Focused tests measure the body transfer reduction directly. Actual wire savings require post-rollout measurement. | Scenario | Baseline response-value bytes | Changed response-value bytes | | --- | ---: | ---: | | Received | 263,276 | 131,547 | | Due retry | 263,342 | 131,613 | | Stale processing | 263,291 | 131,563 | | Future retry | 131,692 | 136 | | Active processing | 131,642 | 86 | | Processed | 131,659 | 103 | The benchmark compares response status/body, all stable durable action fields, and the replay body, signature, delivery ID, and provider routing. All six valid scenarios match. It excludes generated row IDs and timestamps. The final measurement records treatment commit `f1ad8ce838700237b56f895cc61a9c6ed0850cd6`, an empty initial worktree status, and the SHA-256 of the actual source. It ran from the server directory to verify that Git paths resolve from the repository root. Local timing ranges overlap and vary with host load. An earlier loaded repeat had a due-retry median of 19.61 ms before and 22.58 ms after, so no consistent latency gain is claimed. The byte reduction is the verified result. Response-value estimates exclude PostgreSQL protocol/TLS overhead and JSONB whitespace. These are local measurements, not a claim about deployed latency or a monthly bill. `pnpm -r typecheck` and `pnpm build` passed on production-change commit `6f4f36bff525494dadda30b49ea6a9a33976e1ab`. Greptile completed with 5/5 and no review threads on that commit. Commit `2abf9b1e50c568b632e7568398a5826a6122e456` changes only the Git test matcher. Commit `f1ad8ce838700237b56f895cc61a9c6ed0850cd6` fixes benchmark provenance after automated review. Commit `ee2a96dad549c72da3132d2b93fd11eb1c45bc31` repairs an existing async UI test. Final commit `994d398d12be65f595b305d1c2f8a61b9d587ff9` extends real PostgreSQL contention coverage after review. Full CI completed successfully on that final head: [workflow run 40537698349](https://github.com/paperclipai/paperclip/actions/runs/40537698349). All 55 successful checks, one neutral review, and two optional skipped checks are terminal; none are pending or failing. Fresh Greptile and Paperclip reviews are both 5/5, with zero unresolved review threads. The production source SHA-256 remains `402f9c66cd030c9ded9f39a3bc72d26c56627752c0950b102adcab15c4512df3`, so the production proof is unchanged. The full local `pnpm test:run` was not completed because this host exhausted embedded PostgreSQL semaphore resources; the focused real PostgreSQL tests used a private disposable container. Current-head remote CI passed the full suite. A direct server typecheck before runner preparation reported missing runner build outputs; supported preparation and the full typecheck then passed. The first complete build found a partial local Copilot closure from the dependency cache. The complete pinned closure was restored and verified, then the complete build passed. The first local test run also found a fixture teardown order error; the teardown was corrected, then all 17 tests passed. Frozen dependency installation failed on the existing master patch metadata. Independent cached dependency directories were cloned, their workspace links were checked, and no dependency manifest or lockfile was changed. The first CI server4 shard failed one existing Git expiration test: Git 2.56 reported `gitfile does not point to a valid repository`, while the test expected the Git 2.55 phrase `not a git repository`. The service returned the correct `inspection_failed` result and preserved the workspace. Separate test-only commit `2abf9b1e50c568b632e7568398a5826a9c6ed0850cd6` accepts either precise diagnostic. The affected assembled test passed locally with real PostgreSQL 17.11 and Git 2.55; the captured CI diagnostic also matches, and unrelated permission errors do not. This was a compatibility failure, not a transient failure or an unchanged rerun. Paperclip Review found that the benchmark recorded its own scratch directory as an uncommitted change. The final benchmark-only repair captures repository status before scratch creation, runs Git commands from the repository root, records the source hash before measurement, and rejects source edits during measurement. A full benchmark run from the server directory then reported clean provenance and passed all six byte/behavior comparisons. The review thread was marked addressed after this proof. The next full CI run found a separate unchanged UI test failure in `ConnectionIntentInteractionBody.test.tsx`: the saved authentication link was still absent after one fixed flush. The same failure reproduced locally: one failed and 50 passed. Separate test-only commit `ee2a96dad549c72da3132d2b93fd11eb1c45bc31` uses the existing bounded `waitForAssertion` helper around the identical destination assertion. The whole file then passed all 51 tests, and UI token gates passed. The no-extra-setup and no-automatic-completion assertions remain. No UI runtime or timeout changed. Paperclip Review also asked for parallel claims from stale-processing and due-failed rows. The contention test now covers those states and received rows. Each case holds the claims until all 12 workers have read the same eligible attempt, then uses real PostgreSQL to assert one replay and exactly one attempt increment. The whole focused suite passed all 19 tests. This coverage change does not modify production code or the measured queries. The review thread was marked addressed after this proof. ## Risks - Payload validation now follows claim commit. If the process stops there, the existing stale-claim recovery path retains the body and retries it. Tests cover fresh and stale processing rows. - A malformed future retry now fails when due instead of being cleared by an unrelated early worker. The claimed-attempt guard prevents invalid settlement from overwriting a later claim. Terminal invalid-payload settlement intentionally retains the baseline result shape (`code` and `retryable: false`, without `attempts`). Changing that result contract is outside this byte reduction; the non-retryable row cannot become due again. - Existing runtime-generation, credential, recovery-source, replay-barrier, redaction, and retry checks remain in place. Existing assembled tests cover these boundaries. - There is no migration, startup job, dependency change, or deployment in this PR. ## Model Used - OpenAI GPT-6 through Codex. The exact runtime model ID and context window are not exposed by this environment. Capabilities used: source analysis, code editing, shell tools, PostgreSQL experiments, and automated tests. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes (The public contract is unchanged; benchmark usage and evidence are documented above.) - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
7a0c9c23cf |
perf(github): reduce inherited review-wait transfer (#17893)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - GitHub review tasks can retain a passive wait after a review is published. > - The wait check reads current requester, task, agent, repository, and publication bindings. > - It also transfers the complete assessment for one commit comparison. > - Large review summaries and findings add repeated database transfer. > - This pull request selects only the typed reviewed commit value. > - The wait keeps its existing checks and transfers less data. ## Linked Issues or Issue Description **What existing behavior does this improve?** The inherited GitHub review-wait check in the server. **Current behavior** The joined query selects the complete assessment JSON. The caller uses only `assessment.reviewedCommit`. **Proposed behavior** Select `assessment->'reviewedCommit'`. Keep its JSON type and the strict comparison with the head SHA. Keep all joins, filters, ordering, and publication checks. **Reason and benefit** The synthetic PostgreSQL 17 fixture reduces one result DataRow from 6,874 bytes to 423 bytes (93.85%). This measurement excludes response metadata, TLS, and TCP. It does not predict a monthly bill or deployed savings. **Breaking changes** None. No public API, stored data, schema, or execution policy changes. **Additional context** Related work: #17562 narrowed a different external-review candidate query. This PR changes the inherited published-review lookup. The duplicate search found no matching open change. ## What Changed - Select the commit as JSON, rather than text, to preserve strict comparison for legacy numeric, object, array, missing, and null values. - Add real PostgreSQL regressions for narrow responses, unchanged stored assessments, revocation, tenant and agent bindings, and a newer queued head. - Add a synthetic benchmark that captures the real baseline and treatment SQL, compares outputs, and measures raw DataRow bytes. - Document the read contract and benchmark command. ## Verification - Baseline: `663fe667d66d3863d0919a8724f1d99500feef0b`. - `vitest run server/src/__tests__/github-review-wait.test.ts`: 14 tests passed on disposable PostgreSQL 17.11. - Restoring the baseline query makes the response-size regression fail: 8,823 column-value bytes exceed the 1,000-byte bound. Restoring treatment passes. - Benchmark: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-github-review-wait.ts --baseline-ref=663fe667d66d3863d0919a8724f1d99500feef0b`. - The benchmark compares every unchanged SQL column and full service output. It also compares 15 JSON cases and live connection revocation. - Two warmups and seven alternating samples had overlapping latency ranges. No latency improvement is claimed. - `pnpm -r typecheck` and `pnpm build` passed locally. - Twenty existing published-review consumer regressions passed on PostgreSQL 17.11. These cover continued review, a newer head, publication state, and a connection-disable race at final commit. - Final-head focused tests passed again after the benchmark-only review fix. - Final benchmark treatment: `d94410a6e6bb83e727afdf09cff450a91a89edd6`, with an empty source diff. DataRow bytes remained 6,874 to 423. Warm medians were 3.11 ms before and 3.43 ms after, with overlapping ranges. This proves byte reduction, not a latency improvement. - A provenance check with one staged and one unstaged source edit confirmed that the benchmark records both from the documented server working directory. Those temporary edits were removed. - All final-head CI checks passed for `d94410a6e6bb83e727afdf09cff450a91a89edd6`. Greptile scored this exact commit 5/5. There are zero unresolved review threads or recommendations. The independent Paperclip review also scored this commit 5/5. - The first local embedded PostgreSQL attempt failed before assertions because the host exhausted System V semaphores. A direct `initdb` reproduced `could not create semaphores: No space left on device`. The passing focused proof used an owned disposable Docker PostgreSQL 17 container. `pnpm test:run` was not completed locally because of that host limit. The full final-head CI suite passed. - The final CI aggregate briefly reported `in_progress` after all of its steps and the workflow had completed successfully. One rerun of only that aggregate produced a terminal success. The test matrix was not rerun. - The initial frozen install hit the existing upstream patch/lock configuration mismatch. A cached non-frozen local install succeeded. The generated lock change is excluded from this PR. ## Risks The selected commit must retain its JSON type. Using `->>` could turn numeric `1` into text and accept a malformed assessment against head `1`. The regression covers this case. The complete assessment remains available for actual review and publication. Existing fresh authority checks remain in place. No migration is required. ## Model Used OpenAI GPT-6, accessed through Codex. The exact runtime model ID and context-window size were not exposed in this session. Used repository inspection, code execution, and PostgreSQL validation. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
8548173fc9 |
feat(evals): connect hosted Runner costs and traces (#17889)
## Thinking Path > - Paperclip manages AI agents and measures their behavior with Runner evaluations. > - The cost controller, native tracing, private receipts and portable runtime are on master. > - The maintained live workflow still launches the provider CLI directly. > - Independent workers need one shared selection and disjoint budget allocations. > - Reports must keep unresolved allocations when a worker produces no evidence. > - This PR connects the maintained workflow to the existing controller and private archive. > - Operators can inspect known costs, conservative holds and trace evidence before qualifying paid recurrence. ## Linked Issues or Issue Description Refs #17452, #17469, #17493, #17506, #17523. **Current behavior** The live Runner workflow does not use the merged cost/tracing controller. Attempt-only aggregation cannot reconstruct a campaign when every worker archive is absent. Ordinary Actions artifacts in this public repository are readable by authenticated public GitHub users. **Proposed behavior** Controlled runs prepare one manifest, bind every selected request, enforce disjoint allocations and retain evidence in private storage. The report starts from the full retained manifest. Missing workers keep their allocations. The workflow rejects retries, unsupported providers, changed requests and absent or stale operating bounds before provider work. **Reason and benefit** The maintained workflow needs auditable costs and traces, including failures and missing receipts, before paid recurrence can be qualified. ## What Changed - Add controlled mode as the manual default and the only scheduled mode. Legacy mode remains explicit for other manual provider qualifications. - Prepare the full request selection and one shared manifest. Bind current run, first attempt, code, definitions, executable hashes and exact profile allocations. - Verify the copied controller and manifest before provider credentials. Require the controlled target to equal the trusted workflow revision before private archive access. - Reuse private archive retention for the controller policy, cell attempts and canonical report. AWS credentials remain transfer-step outputs. Controlled raw evidence is never uploaded as an Actions artifact. - Replay a retained manifest with partial or empty worker evidence. Count operating holds once and preserve known partial charges. - Pin the controller to an explicitly reviewed compatible lockfile digest. Reject resolver or byte drift before a frozen, script-free install into a job-local store. Record the lock digest in the runtime receipt and verify it again in private preparation. - Build the portable controller without secrets, private environment access or OIDC. Private preparation verifies that artifact and performs no dependency install. OIDC jobs use no mutable dependency resolution or shared cache. - Add synthetic source/copied-runtime tests, executable workflow admission tests, private archive tests and an operator runbook. ## Verification - Controller, campaign, portable runtime and workflow tests: 104 passed. This includes real aggregation from an isolated reporter without source dependencies, safe diagnostics, driver drift, distinct evidence binding and reviewed-lock drift. Real private Codex/Claude requests also pass through an independently verified copied runtime with fake transports. - Private transport and workflow contracts: 35 passed. Executable/static workflow guards: 13 passed. - Paired private report tests: 166 passed; Evalbook discovery: 7 passed. - `pnpm -r typecheck` and `pnpm build`: passed. - `pnpm test:e2e:runner:typecheck`, `actionlint .github/workflows/runner-protocol-live-evals.yml` and `git diff --check`: passed. - `pnpm test:run` hit repeated local embedded PostgreSQL initialization failures in unrelated DB-dependent server tests. Its owned process tree was stopped after preserving the log. This is a local verification limitation; current-head CI remains required. - An earlier head hit one unchanged Rust cleanup assertion. The exact test passed locally and the final head passed the full Rust lane; the initial failure log is retained. - On the final head, one unchanged chat idle-queue assertion failed once in CI. Its single scoped retry passed. All test jobs and the final aggregate gate passed on commit `2e328fa53dd296174d59d19f14f068aeeb0b82d8` (CI run 40535692048, attempt 2). - No paid model, Braintrust or cloud calls. No live variables, secrets, schedules or resources changed. ## Risks - Controlled mode fails closed until a reviewed `RUNNER_PROTOCOL_EVAL_CONTROLLER_POLICY_JSON` and a compatible private eval SHA exist. `all` does not silently become a supported subset. - The reviewed policy must cover setup, model requests, trace/gateway processing, archive transfer and retained storage. Hash binding does not establish vendor billing bounds. This PR does not enable recurrence or release historical holds. - The controlled policy must pin `controllerLockSha256`; it never accepts a resolver-generated digest automatically. No lockfile is committed. A changed compatible lock needs review. - Legacy browser verification now requires the committed frozen lockfile. A stale lockfile fails closed; controlled preparation has no dependency install. - Workflow cancellation can prevent the final report. The full manifest is retained before provider work and remains the starting point for reconciliation. - The workflow exceeds 1,000 lines because the shared shard anchor, dependencies, environment expressions and credential boundaries belong to one declarative job graph. Request preparation, accounting, archive safety and runtime verification stay in separately tested modules. ## Model Used OpenAI Codex, GPT-6-based coding agent with tool use and code execution. The host does not expose an exact deployment model ID or context-window size. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run the relevant tests locally and they pass; the broad local suite limitation is documented above - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green on the final head - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups; the secondary reviewer is also 5/5 - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
3be54534c2 |
perf(server): decode workspace run context keys once (#17898)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Managed Git workspaces retain task and run obligations until cleanup is safe. > - Activity, expiration, retention and rehome share a run association query. > - That query repeatedly extracted three keys from wide historical JSON contexts. > - This pull request decodes those keys together and projects the six fields its consumers use. > - Native bindings, cleanup vetoes, lifecycle locks and fresh deletion checks stay in place. > - The benefit is less database work while preserving the same cleanup decisions. ## Linked Issues or Issue Description **What happened?** Workspace activity and eligibility queries repeatedly decompress large run contexts. An OR predicate extracts issue, task and workspace compatibility keys separately. The run scope also materializes columns its lifecycle consumers do not read. **Expected behavior** Read the same relevant runs and preserve all deletion and rehome vetoes with less repeated JSON work. Keep existing native, historical and pointer bindings independent. **Steps to reproduce** Run the synthetic PostgreSQL comparison below against baseline `db70c5acd92f50fa37fb4f0dea3da0e7685089a2`. Repeat the small, large-tree and combined controls documented in the PR. Related work: #17464 preserves workspace descendants. #17877 reconstructs native workspace obligations. This change preserves both contracts and adds no scheduling change. ## What Changed - Decode the three legacy context keys together with `jsonb_to_record`. A strict root-object guard preserves scalar, array and null behavior. Independent native and pointer bindings still match. - Materialize the full textual tree-ID set once. This avoids inflated recursive estimates triggering JIT compilation for the correlated decoder. - Project the six run lifecycle columns used by activity, expiration, orphan retention and rehome. - Add binding, lease, malformed-shape and fresh-read regression cases. Add a reproducible four-control benchmark and detailed evidence documentation. ## Verification - Full repository typecheck passed on the integrated final source. - Six relevant suites passed all 209 tests against real PostgreSQL, including the 20 new binding/shape cases, 25 eligibility cases, full Git expiration/restoration, task-directory retention and configuration, and isolated repository rehome. These cover conflicting bindings, company isolation, duplicate admission, pending cleanup, controller leases, non-object contexts and newly active consumers. - PostgreSQL 17.11 comparisons use two warmups and five alternating samples per variant. Each sample runs eight sequential workspace queries. Every complete caller result and six-column relevant-run multiset matches. - Stress fixture: 4,813 runs and about 332 MB of run storage, with unique wide contexts. Last-activity median: 1,071.54 to 743.45 ms per eight workspaces, 1.44 times faster. Closed eligibility: 814.29 to 713.86 ms, 1.14 times faster. Idle eligibility: 721.46 to 481.69 ms, 1.50 times faster. - Small-history control improves 1.23–1.36 times. The 250-descendant stress control improves 1.24–1.30 times. Combined small-history/large-tree activity and idle checks improve 1.19–1.22 times. Combined closed eligibility is neutral: 61.11 to 62.82 ms per eight, about 0.21 ms extra per workspace. - Stress last-activity ranges do not overlap: 920.27–1,111.22 ms before and 689.14–880.40 ms after. Closed-eligibility ranges overlap; its smaller median gain is directional. - Last-activity buffer accesses fall from 113,202 to 38,081. Eligibility run width falls from about 1,414 to 57 planner bytes. All final plans retain one history scan with no JIT. - Reproduce: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-workspace-activity.ts --baseline-ref=db70c5acd92f50fa37fb4f0dea3da0e7685089a2 --docker-postgres=17 --reject-new-jit --output-dir=/tmp/workspace-activity-stress`. - Repeat with `--small`, `--descendants=250`, and both flags, using distinct output directories. Each run creates and removes its own synthetic database. - Full CI passed on `bc4edd700f983e22f093d0a01e2f4c795075b982`: 54 successful checks, one neutral review and two expected Storybook skips, plus successful Snyk status. Greptile is 5/5 on this exact head. There are zero unresolved threads and the merge state is clean. [CI run](https://github.com/paperclipai/paperclip/actions/runs/40530494291). - Full repository typecheck and build passed on the integrated final source. The local full test command was stopped after native PostgreSQL initialization failed repeatedly, producing 16 failures and skipped database suites. It is incomplete, not a pass. Focused real-database suites used disposable Docker PostgreSQL. Current-head CI passed all full gates separately. ## Risks - Cleanup safety depends on complete run association. Tests compare every retained run field and exercise downstream vetoes, not only row counts. - PostgreSQL plan estimates matter. Rejected per-key index probes were up to 56 times slower on a small-history/large-tree control. An ANY variant also regressed. The final shape passes all four controls under default JIT settings. - The final combined eligibility control is neutral. Synthetic SQL timings do not establish deployed latency, provider behavior or actual compute sleep. - Review suggested a CASE/jsonb_typeof root guard as an optional alternative. The strict JSONPath guard is retained because its scalar, nested-array and independent-native behavior passes real PostgreSQL tests and all four measured controls. No correctness defect was identified. A replacement needs its own comparative measurements. - Future consumers that need another run column must extend this explicit projection. - No migration, index, startup job, polling change or deployment is included. ## Model Used OpenAI Codex, configured model `gpt-6.1-sol` with xhigh reasoning, tool use and code execution. Independent blast-radius review also used `gpt-6-astra` and `gpt-6.1-sol`. The runtime context window was not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
475f7df2e7 |
perf(server): prefilter chat milestone run candidates (#17892)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Chat publications connect a run to its authorized originating conversation. > - The milestone selector evaluates detailed lineage and final-answer predicates. > - A local plan evaluated these predicates across many unrelated run and conversation pairs. > - This pull request materializes recent runs that already match an active or waiting conversation. > - The existing query still makes each publication decision. > - The benefit is less JSON work when most run history is unrelated to active chat. ## Linked Issues or Issue Description **What happened?** The milestone candidate query repeatedly reads large run context values for unrelated conversations. A synthetic baseline discarded 60,962 run and conversation pairs through the complex join filter. **Expected behavior** Apply the existing same-company conversation binding before expensive lineage checks. Preserve every candidate column, order, publication decision and continuation rule. **Steps to reproduce** Run the two synthetic PostgreSQL comparisons below against baseline `db70c5acd92f50fa37fb4f0dea3da0e7685089a2`. They capture complete SQL from the actual service implementations. Related work: #17606 changes chat lifecycle scheduling. #15802 deferred milestone cadence. This SQL-only change leaves scheduling, startup reconciliation and recovery intact. It has no dependency on those PRs. ## What Changed - Materialize recent runs with an active or waiting same-company conversation, using the unchanged native and legacy binding expression. - Retain every outer lineage, continuation, final-response, privacy, cursor, ordering and idempotency check. - Add a real PostgreSQL historical-binding regression and selective/dense performance controls. - Document the measured gain and the dense-history tradeoff. ## Verification - Dedicated binding, publication reconciliation and app coordination suites passed all 27 tests across three files with real PostgreSQL where required. - The selective fixture has 2,615 runs, 55 conversation issues and about 192 KiB of pseudorandom context padding per run. Five alternating measured samples follow two warmups on PostgreSQL 17.11. - Final-harness selective milestone median: 1,495.85 ms before, 130.96 ms after, 11.42 times faster. Ranges: 1,483–1,716 ms and 125.59–135.85 ms. All 60 complete ordered rows match, including 15 explicit historical-binding cases. - Dense control: 736.71 ms before, 809.08 ms after, with overlapping ranges. All 200 complete ordered rows match. This is inconclusive performance evidence with possible materialization overhead. - The unchanged native progress control is neutral in both fixtures. It returns 41 equal complete ordered rows. - The added large-window control has 12,215 runs, including 12,000 recent native history rows. A caller limit of one produces the real service page size of 25. First-page medians 762.65 to 800.63 ms overlap and are inconclusive. The 90% cursor improves 234.84 to 54.36 ms; the 99.99% tail improves 352.10 to 12.88 ms; the finished cursor improves 340.00 to 58.58 ms. Complete ordered results match: 25, 25, seven and zero rows. Same-run later-conversation rows remain present. - Large-window CTE materialization itself takes about 4.6–10 ms with zero temp spill. This does not establish a materialization regression or justify a larger production query rewrite. - Reproduce the added window control with `--cursor-window` and a distinct report path. - Reproduce selective: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-chat-milestones.ts --baseline-ref=db70c5acd92f50fa37fb4f0dea3da0e7685089a2 --docker-postgres=17 --report=/tmp/chat-milestones.json`. - Add `--dense-chat` and a distinct report path for the dense control. The harness applies normal migrations to a disposable container and uses no hosted database. - The existing real-service test at `server/src/__tests__/chat-channels.integration.test.ts:34584` calls enqueue with limit=1 four times, crosses 30 resolver-rejected candidates, expects admissions [1,1,1,0], and asserts the exact three publication keys plus internal-run exclusion. It passed and supplies the full composition/publication-set proof requested in review. - Assembled chat integration passed 37 selected tests. Together with the 27 focused tests, 64 checks passed. The selector covers bounded draining, exact continuations, private summaries, final-answer precedence, lifecycle idempotency and native progress races. - Final head `1b3a6f50c9e7ec712dcbaf82c7fc3df141ec3f28`: [full CI passed](https://github.com/paperclipai/paperclip/actions/runs/40534934737), with 55 successful checks, two skipped and one neutral. Greptile is 5/5 on this head; zero unresolved review threads. Paperclip Review is 4/5, neutral, explicitly with no must-fix correctness or security defect. Its remaining payload-memory/dense-overhead caution is retained here as a measured residual tradeoff; an identity-only projection needs production evidence and separate comparison. - After the Drizzle metadata correction, both generated service queries and every parameter remain byte-identical to the selective and dense measured versions. The real PostgreSQL binding regression passed again. - Full repository typecheck and build passed on the integrated final source. The local full test command was stopped after native PostgreSQL initialization failed repeatedly, producing 16 failures and skipped database suites. It is incomplete, not a pass. Focused real-database suites used disposable Docker PostgreSQL. Current-head CI runs the full gates separately. ## Risks - A materialized source can add overhead when most recent runs match active conversations. The dense control does not establish a gain. - Historical JSON comparisons retain string and whitespace semantics. Test fixtures disable the normalization trigger only while inserting retained historical shapes in disposable databases. - Synthetic SQL measurements exclude provider delivery and do not establish deployed latency, actual compute sleep or fleet-wide cost reduction. - No schema migration, index, startup job, polling change or deployment is included. ## Model Used OpenAI Codex, configured model `gpt-6.1-sol` with xhigh reasoning, tool use and code execution. Independent blast-radius review also used `gpt-6-astra` and `gpt-6.1-sol`. The runtime context window was not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
a26d217e94 |
fix(server): coalesce Cloud membership synchronization writes (#17890)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Cloud requests synchronize a human user's profile and company membership. > - Concurrent requests repeated the same writes before the debounce could record success. > - Upserts also updated unchanged rows and caused avoidable tuple work. > - This pull request shares identical pending writes and guards unchanged updates. > - Each request still reads current authority. Bounded profile lookup also reaches older memberships. > - The benefit is less database work during request bursts with the same actor results. ## Linked Issues or Issue Description **What happened?** A burst of 20 trusted Cloud requests issued 20 membership upserts and 40 profile plus membership tuple updates for one unchanged user. The completed-write debounce did not share in-flight work. **Expected behavior** Identical in-flight requests share persistence work. Distinct contexts retain their order. Failures remain retryable. Each request resolves its own current company and role authority. **Steps to reproduce** Run the synthetic PostgreSQL comparison below at baseline `db70c5acd92f50fa37fb4f0dea3da0e7685089a2`. The harness imports the real resolvers and counts actual writes with database triggers. Related work: #17663 covers adjacent Cloud authorization and projection behavior. This change preserves fresh imported memberships, owner-elevation settings and stale administrator removal. ## What Changed - Serialize synchronization per database client and user. Share only the last identical queued context. Preserve A/B/A transitions and retry after partial failure. - Guard user and membership updates with null-safe value comparisons. Retain default-grant repair when `RETURNING` is empty. - Keep authority reads outside the write debounce. - Resolve exact principal IDs through the existing index, then use at most one company-scoped alias query. This prevents both the old 200-row profile cutoff and repeated membership-page scans. Exact IDs take precedence over namesake aliases. - Add concurrency tests, PostgreSQL tests, a reproducible benchmark and behavior documentation. - Accept both Git 2.55 and 2.56 diagnostic messages in an existing workspace safety assertion, after full CI exposed that portability failure. No expiration behavior changes. ## Verification - Five focused suites passed, 63 tests total. These include Cloud actor concurrency, transient database retry, company provisioning, real PostgreSQL synchronization and the user profile route. - The auth-to-profile regression returned 404 before the lookup repair and 200 after it. The final fixture has 601 memberships and tests raw and normalized non-UUID IDs, missing slugs, alias ordering and exact-ID priority. A counterfactual with the old paging repair performs four membership queries for a missing slug; the final code performs two, each returning at most one row. Cross-company access returns 403 before lookup. - Local PostgreSQL 17.11 benchmark uses two warmup pairs and five alternating measured pairs. Each pair starts a fresh debounce scope. All 20 complete actor results match on every sample. - Burst median: 51.07 ms before, 25.71 ms after. Ranges: 47.21–63.25 ms and 24.17–29.32 ms. Membership statements: 20 to one. Tuple updates: 40 to zero. - Reproduce: `pnpm --filter @paperclipai/server exec tsx ../scripts/benchmark-cloud-membership-sync.ts --baseline-ref=db70c5acd92f50fa37fb4f0dea3da0e7685089a2 --docker-postgres=17`. - Real PostgreSQL tests compare all four alias fields with the route's own JavaScript canonical slug output across 76 fixtures. Exhaustive Unicode lowercase inspection validates the special SQL mappings. - A real SQL_ASCII database reproduces the previous en-dash alias 404. The final whole-character replacements precede ASCII-only translation. Four alias fields and ten multibyte edge cases pass, including rejection of a false euro-to-k alias. The disposable database is migrated normally, then removed. - Server TypeScript passed with no diagnostics. - PostgreSQL tests used the normal migrations with a temporary Docker startup override because this host's native PostgreSQL semaphore allocation is exhausted. No tracked test infrastructure changed. - The first full CI run had an unrelated CLI company-import/export health-start timeout. Review also identified repeated profile scans; this follow-up repairs that finding. A later run passed 3,882 tests in server shard 4/12 and failed one existing assertion because Git 2.56 changed its error wording for a missing gitfile target. The safety behavior still refused expiration and preserved the workspace. The assertion now accepts both valid messages; the actual Git/PostgreSQL regression passes locally. Final head `3a644e600ab0ae7315f9828b1ea9dffdef407e4a` passed [full CI](https://github.com/paperclipai/paperclip/actions/runs/40537153522), including the previously failing shard. There are 55 successful checks/statuses, two skipped and one neutral. Fresh Greptile and Paperclip reviews are both 5/5 on this exact head, with zero unresolved threads. GitHub reports mergeable and CLEAN. - Full repository typecheck and build passed on the integrated final source. The local full test command was stopped after native PostgreSQL initialization failed repeatedly, producing 16 failures and skipped database suites. It is incomplete, not a pass. Focused real-database suites used disposable Docker PostgreSQL. Current-head CI runs the full gates separately. ## Risks - Coalescing is local to one process and database client. Other replicas can still take conflict locks. - `updated_at` now records actual profile and membership changes. Access lists sorted by it can show a different order. Profile resolution no longer depends on the first 200 entries. Exact principal IDs take precedence over namesake aliases. The single alias fallback still scans company candidates; no constant-CPU claim is made. - Default role grants retain their existing additive semantics. This change does not define a new demotion policy. - Local synthetic service timings exclude HTTP and network latency. No hosted deployment or production speedup is claimed. - No migration, startup job, polling change or deployment is included. ## Model Used OpenAI Codex, configured model `gpt-6.1-sol` with xhigh reasoning, tool use and code execution. Independent blast-radius review also used `gpt-6-astra` and `gpt-6.1-sol`. The runtime context window was not exposed. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
aec86c794b |
fix(runner): preserve native work through controller outages (#17594)
Keep remote native work under its authenticated connection lease through controller outages. Hold durable ACPX tools through reconnection, align the native Claude HTTP MCP client budget, and wait for the recovered handler before dispatch. Preserve Stop, cancellation, exact identity and artifact checks. Verified with real-process regressions and a real Claude Opus 5.5 Daytona workflow: 125 seconds offline during useful work, original process and lease retained, no failed native tool calls, queued follow-up completed, independently verified ZIP and committed workspace receipts. All current-head CI gates pass and Greptile is 5/5 with all findings resolved. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
4d90634f74 |
feat(connections): add 1Password service-account connector (#17249)
## Thinking Path > - Paperclip lets people manage AI agents and govern their access to outside services. > - Apps connections keep provider credentials in the instance vault and mediate every action through grants, policies, and audit. > - Agents currently have no Apps connection for reading or managing 1Password items with an operator-supplied service-account token. > - 1Password supports this through its official SDK; its documentation MCP server does not expose vault operations. > - This pull request adds a one-field service-account connection and registers eight reviewed SDK actions through the existing native-provider catalog. > - Operators can select vault permissions in 1Password and agent/action access in Paperclip without installing a CLI or handing durable credentials to agents. ## Linked Issues or Issue Description Related: #16144 adds 1Password as Paperclip's secret-storage backend. This PR adds the separate Apps/tool-access flow with a per-connection credential; it does not change secret providers or use a server-wide ambient token. **Subsystem affected** Apps catalog, native tool-provider registry, and shared connection setup. **Problem or motivation** People need to paste their own `OP_SERVICE_ACCOUNT_TOKEN`, find official service-account setup instructions, and authorize agents to use the selected vaults through ordinary Paperclip governance. **Proposed solution** Add `1password/service-account`, with a password field labeled **Paste your 1Password Service Account token** and a link to [1Password's creation instructions](https://www.1password.dev/service-accounts/get-started#create-a-service-account). Use the pinned official `@1password/sdk` 0.5.0 for vault/item listing, item and secret-reference reads, and item create/update/archive/delete. **Alternatives considered** A remote MCP card would not provide vault access. A local CLI or a third-party MCP package adds installation and credential surfaces that the official SDK avoids. ## What Changed - Added the source and generated catalog definition, permission review, official light/dark artwork, documentation, and capability-map references. - Added a provider-owned SDK bridge with strict schemas, account-domain validation, contract hashes, fixed errors, bounded output, and disposable workers with cancellation/deadlines. - Registered the provider through master's native-provider catalog; existing setup, health, discovery, gateway, policy, and audit owners remain in place. - Labeled the single password field **Paste your 1Password Service Account token**. The shared formatter preserves explicit credential prompts; ordinary key labels retain their existing presentation. - Added credential-ownership, gateway read/write, policy/revocation, worker, setup UI, and interactive Storybook coverage. ## Verification - Final review follow-up: added two deterministic worker-entry tests for nested exact-token redaction and oversized-output rejection. Removing each guard caused exactly its regression test to fail. The focused provider suite passed **15 tests**, and the official server typecheck passed after rebuilding stale workspace SDK artifacts. Current head `569cb7f9bd57db791fb13a6034ce69bf50878cb1` passed the full [CI run](https://github.com/paperclipai/paperclip/actions/runs/40016166260), including typecheck, build, unit/integration tests, and all eight browser shards. The required `ci / verify` and `ci / e2e` gates passed. [Greptile](https://github.com/paperclipai/paperclip/pull/17249#issuecomment-6353731805) and [Paperclip Review](https://github.com/paperclipai/paperclip/pull/17249#issuecomment-6372165895) each rated this exact head **5/5**, with no open findings. Paperclip Review completed with a neutral check because its scoped reviewer could not inspect CI; the full CI result was verified separately. Other neutral checks and intentionally skipped Storybook jobs contain no actionable findings. - CI recovery on the same commit: the first attempt lost six jobs. Five logged simultaneous runner shutdowns at 18:49 UTC on 2026-10-23; the sixth stopped logging before a test result and its root cause was not established. No assertion failure appeared in the retrieved logs. All six jobs and the aggregate gates passed one failed-job retry. [CI run](https://github.com/paperclipai/paperclip/actions/runs/40005754936). - Copy follow-up: the catalog, provider, and setup UI suites passed **275 tests**. The governed gateway lifecycle passed **3 tests** (one live-only test skipped), including create, update, archive, and delete. UI token gates and map checks passed. The branch was rebased onto `9c442cf48e` to resolve additive catalog conflicts while retaining the Vercel and Supabase entries. The existing live write proof below remains applicable; this follow-up changes presentation only. - Five affected suites, including the existing Daytona consumer, passed **290 tests** (one opt-in live test skipped in fixture mode). Full `pnpm -r typecheck` and `pnpm build` passed. The local broad-suite attempt could not complete because PostgreSQL semaphore exhaustion prevented a competing integration suite from starting; the complete affected-file retry passed after disposable-process cleanup. The prior commit `f208a8aff6dc28859cc3f343497e18d85f124579` passed the complete CI matrix. All CI build, typecheck, test, and browser gates also passed on the earlier commit `9e2f3d3140368a08da43b004b2e8d3a0f4c0fa4d`. The separate Paperclip Review reported a worker-result regression-test gap on that head; the follow-up above addresses it. Its review thread is resolved, and both fresh current-head reviews are clean as recorded above. - Token gates, module boundaries, brand validation, map checks, and Storybook catalog checks passed. - Direct SDK create/archive/delete and rejected-call coverage passed (13 provider tests); the Storybook Permissions/Test play completed successfully in Chromium. Server and UI typechecks passed after those fixes. Greptile passed prior commit `9e2f3d3140368a08da43b004b2e8d3a0f4c0fa4d` at **5/5**. Both reviewers also passed the worker-test follow-up at **5/5** as recorded above. - Live 1Password proof on 2026-10-22: connect, eight-action discovery, vault listing through an actual run-scoped gateway, refresh, reconnect preserving identity, another read, removal, and denial through the retained session passed. Credential absence was asserted in responses/audit; an exact-token scan of inspected logs and changed files passed. - Real browser: Connectors → 1Password → token entry → personal access limited to one disposable agent → completion → Permissions → Test `onepassword_list_vaults` succeeded. The disposable connection was removed afterward. - Compiled-worker smoke initially exposed inherited Node launch flags. Workers now use explicit empty `execArgv`; the compiled worker then authenticated successfully and listed one vault. - Live write proof on **2026-10-23**: the replacement Keychain token passed all eight actions through the actual run-scoped gateway for both personal and shared credentials. The test created, listed, read, updated, resolved a secret from, archived, and deleted two synthetic items. Both were absent from subsequent item lists. Refresh/reconnect, identity preservation, removal, retained-session denial, and credential absence passed. The live suite passed **3 tests** (one fixture-only test skipped) in 116.65 seconds. The expanded fixture suite passed **3 tests** (one live-only test skipped); the 13 provider tests and server typecheck passed. An exact-token scan passed over 64 source/log files. - Failure history: the original token denied creation on 2026-10-22 and created no item. The first 2026-10-23 attempt stopped before provider calls because host PostgreSQL semaphore limits prevented test-database startup. A temporary ten-connection test profile allowed the complete retry to pass without changing shared services or host limits. A new delete-result assertion also exposed an outdated fixture response (`success` instead of `deleted`); the fixture was corrected and the complete fixture suite passed. Official contract sources: [service accounts](https://www.1password.dev/service-accounts/get-started), [SDKs](https://www.1password.dev/sdks), [item management](https://www.1password.dev/sdks/manage-items), [official JavaScript SDK](https://github.com/1Password/onepassword-sdk-js). ## Risks - Authorized item/secret reads intentionally return secret values to the agent and ordinary tool-result history. Select trusted agents and use restricted service-account vaults; the connection token itself is redacted. - New connections retain existing Paperclip action defaults, including Allowed writes/destructive actions for selected agents. 1Password still enforces the token's permissions. - Service-account access cannot be changed in place. Create a replacement account and reconnect to change vault permissions; read-only keys can connect but cannot write. - SDK authentication happens per action in a disposable worker. Calls have a 30-second maximum; cancellation stops local work, while a mutation already accepted by 1Password can still finish. Writes are never automatically retried. - Additive dependency and catalog change; no database migration. Disable/remove the connection to withdraw access, or remove the provider from catalog visibility for rollback. ## Model Used OpenAI Codex, based on GPT-6, with code editing, shell/browser tools, and live provider verification. A more specific runtime model ID and context-window limit were not exposed in this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
9c442cf48e |
feat(connections): deploy and configure Supabase and Vercel (#17217)
Add governed Supabase and Vercel deployment/configuration connectors with vaulted credentials, agent access policies, approvals, audit, and bounded provider operations. Deliver reviewed REST actions to native agents and guard Vercel preview intent on empty projects. Preserve secret-safe request identity and durable retry behavior. Verified by current-head CI, fresh Greptile 5/5 review, and real-agent deployment/configuration walkthroughs on disposable provider resources. Test resources and temporary provider tokens were removed after verification. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
||
|
DottaandPaperclip
|
3bd709df8b |
Operationalize Product E2E accounting, tracing, and scheduling (#17009)
## Thinking Path > - Paperclip manages agents and the work they perform. > - Product E2E tests exercise real agents, tools, setup and cleanup. > - Repeated runs need retained costs and traces, including failed attempts. > - A passing retry must not hide earlier spending or feedback. > - Recurring work also needs admission limits, private reports and failure detection. > - This change supplies those controls while keeping paid activation behind existing gates. ## Linked Issues or Issue Description Refs #16996 and #17158. This is the consolidated landing of the reviewed eval operations stack: #17019, #17024, #17025, #17027, #17030, #17033, #17035, #17038, #17043, #17044, #17048, #17050, #17051, #17092, #17096, #17100, #17107, #17109, #17112, #17115, #17117, #17121, #17123, #17127, #17130, #17133, #17134, #17137, #17140, #17143, #17154. The repository permits squash merges only. A local rehearsal showed that squashing each slice into master causes repeated conflicts in dependent slices. This PR preserves their commits and lands their combined implementation in one reviewed integration. Superseded slices will close after this merge. ## What Changed - Retain every attempt's cost receipt, including setup, failures, interruptions and retries. Separate reported amounts, estimates, missing costs and reserved capacity in reports and dashboards. - Add bounded request admission, exact model pricecards, gateway error accounting, cache policy and reconciliation evidence. Refuse unsupported routes and requests that cannot fit the allowance. - Integrate Braintrust gateway traces and OTLP correlation, with secret filtering and experiment imports. Record phase timing, including setup, feedback capture and cleanup. - Collect complaint and suggestion evidence after each attempt, including recovery after failures. Retain explicit review dispositions independently from the product grade. - Add gated daily and weekly selections, measured forecasts, setup/readiness checks, wall-time limits and retained monitor notifications. Include declared non-model overhead in admission while reporting the existing subscription separately. - Keep recurring history private, bound uploads and retention, and use isolated archive environments. Keep public publication an explicit operation. Report and publication jobs install only the committed lockfile, with scripts disabled and isolated stores; stale pins fail closed. - Integrate current master, preserve its Product E2E journey billing and privacy checks, and test that shared setup receipts do not double-count estimated agent usage. Preserve upstream fixture repairs. ## Verification - Complete integration baseline `0649a5ce7544ab20825c2183d2f421157db49f24` includes master `a3e1d915a7110a4e115b46396a3f79e6dfb4d05e` and every original stack head as ancestors. - On that integration baseline, `pnpm test:e2e:runner:unit`: 4,012 Vitest tests passed, two skipped, across 193 files; all 137 Node source-contract tests passed. - Focused billing and Product E2E journey contract checks: 24 tests passed. They cover estimates, unknown receipts, single-owner setup costs, browser-only execution and private publication guards. - Monitor and workflow wiring: all 96 Node tests passed. `node scripts/check-token-gates.mjs`: all gates clean. - `pnpm test:e2e:runner:typecheck` passes after the standard Runner TypeScript build refreshed local declarations. The first attempt rejected the new master `runner_prepared_launch` union member because this checkout retained an older generated declaration; the source union already contained that member. - Current head `492b0708a1c531a023d9ac10f0f3dc6383cc1754` removes dependency re-resolution from archive-authorized reporting jobs after the security scan flagged it. All 15 workflow-security tests pass after updating the old two-install assertion. - Hosted [CI run 39965460180](https://github.com/paperclipai/paperclip/actions/runs/39965460180) passes on the current repaired head, including typecheck, build, general tests, Runner checks, browser shards and canary dry run. Greptile is 5/5 on this exact head with no unresolved threads. The security scan passes after the fixed lockfile-resolution thread was resolved. - Paperclip Review returned a terminal neutral check but an incomplete 0-file assessment on the repaired head: its platform blocked both diff responses as prompt-injection content. This is not a successful code review. No bypass or fabricated pass was attempted; current-head Greptile, security and hosted CI are independent gates. - No paid test, reservation release or recurring-run activation accompanies this landing. Live route qualification and invoice reconciliation remain separate evidence. <details> <summary>Retained earlier validation failures</summary> Earlier broad local runs had two 15-second test timeouts (CLI poll cadence and Retry validation), one embedded-database setup timeout, and later PostgreSQL semaphore exhaustion. The exact timed-out tests passed separately without larger timeouts; the broad-run causes were not established. Those local failures remain distinct from subsequent hosted CI results. The independent #17158 fix records its Muse fixture collision, its UUID correction and hosted verification. Local Muse execution was also blocked by macOS embedded-PostgreSQL startup. No historical failure is treated as a successful paid qualification. </details> ## Risks - Admission is a local control, not a provider invoice cap. Missing provider receipts or resource charges retain unknown amounts and holds; estimates do not become billed dollars. - The bounded path admits only documented local suites and provider routes. Ordinary legacy or separate Runner protocol campaigns have different controls. Merging this code does not qualify additional routes. - Recurrence still needs the existing enable gates and current qualification/allowance evidence. Archive permissions and resources must match the documented private-environment contract. - Legacy reports may lack complete attempt or phase evidence. Reports preserve that limitation and do not infer zero usage from missing data. ## Model Used OpenAI GPT-6 through Codex, with reasoning, repository tools and code execution. The exact deployment variant and context-window size are not exposed by this session. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass (targeted checks; full-suite host limitation disclosed above) - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [x] All Paperclip CI gates are green - [x] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing> |
||
|
Devin Foley
|
059ae49cab |
feat(connections): add Unikraft Cloud connection (#17139)
## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work. > - Apps connections give agents governed access to external services through vault credentials, grants, action policies, the tool gateway, and audit. > - Unikraft Cloud runs unikernel micro-VMs, services, and volumes across regional metros, and it offers a sandbox plugin for commands and files. > - Unikraft is not in the catalog. It has no hosted MCP server, so the generic "Connect your own MCP server" flow cannot reach it. > - Daytona (#16483) added a native REST pattern for this case: a catalog generated from the provider's published OpenAPI specifications and executed through the gateway. > - This pull request adds Unikraft on that pattern and moves the shared native-provider code into one registry, so the shared connection code does not name each provider. > - The benefit is governed agent access to all published Unikraft actions, with the token kept in the vault and sent only to published Unikraft hosts. ## Linked Issues or Issue Description Refs #16483 **Problem or motivation** Teams that run workloads on Unikraft Cloud want their agents to list instances, read logs, deploy and scale services, and run commands in sandboxes. Unikraft is not in the Apps catalog. Unikraft has no hosted MCP server, so operators cannot connect it through the remote MCP flow. **Proposed solution** Add a native Unikraft connection. Generate its actions from Unikraft's official OpenAPI specifications for the platform API and the sandbox plugin. Pin the specifications to a reviewed commit. The operator connects an API token. The token stays in the vault and goes only to the five published metro hosts. Every action takes an optional metro. Actions without a metro use the connection's default metro, which the operator can change under Advanced. **Alternatives considered** A catalog-only AppDefinition was not possible because no hosted MCP endpoint exists. Wrapping the `unikraft` CLI was rejected because it would expose host filesystem behavior and CLI login state. A second copy of the Daytona bridge was rejected; the shared transport and catalog contract are now one module with a provider registry. **Roadmap alignment** This extends the existing Apps and Connections provider coverage. It uses the existing credential, grant, policy, gateway, and audit model. It does not overlap planned core work. ## What Changed - Added `scripts/generate-unikraft-tools.mjs` and its tests. The generator reads the platform and sandbox plugin specifications at commit `5dd107ea1cd8716b83cb86a83ed2a6ee46f99b5b`, resolves shared parameters, records source SHA-256 hashes, and writes `server/src/services/unikraft/operations.generated.json`. - The catalog has 105 actions: 89 platform operations and 16 sandbox plugin operations. The server-sent audit event stream is recorded in the inventory and is not advertised as an action. - Classification: GET operations and sandbox reads and waits are reads. Delete, stop, suspend, detach, unpin, and sandbox command execution, signals, and stdin are destructive. All other operations are writes. - Added `server/src/services/connectors/unikraft/client.ts` and its generated catalog. It validates arguments against the generated schemas, accepts only published metros, rejects path traversal, sends JSON bodies on GET and DELETE as the specification requires, and verifies the token with `GET /v1/users/quotas`. A sandbox wait that times out (documented HTTP 408) returns a still-running result instead of a failure. - Moved the bounded, credential-redacting HTTP sender and the catalog contract helpers from `daytona.ts` to `native-rest-client.ts`. Added `native-provider-catalogs.ts`. `tool-access.ts`, the `tool-access/catalog` and `token-broker` modules, `connectors/tool-policy.ts`, and `tool-gateway.ts` now use this registry instead of Daytona-specific checks. Daytona behavior does not change. - Added the Unikraft definition source, the generated definition, the brand mark and manifest entry, and the permission review (`liveProof: not-run`). - Added `doc/connections/UNIKRAFT.md` with a sequence diagram, endpoints, setup, actions, and limits. Updated the connections README, the permission audit, and both capability maps. - Tests: definition shape, database-backed gateway reads, writes, deletes, and sandbox commands for personal and shared grants, default and per-call metros, rejection of unpublished metros and traversal, Off policies, revocation, stale-contract blocking, credential export denial, schema validation for all actions, pinned-transport tests for GET and DELETE bodies, and a browser setup check. ## Verification - `pnpm exec vitest run packages/shared/src server/src/__tests__/unikraft-connection.test.ts server/src/__tests__/unikraft-http-boundary.test.ts server/src/__tests__/daytona-connection.test.ts server/src/__tests__/daytona-http-boundary.test.ts server/src/__tests__/tool-access-service.test.ts server/src/__tests__/tool-gateway-service.test.ts server/src/__tests__/tool-gateway.test.ts ui/src/pages/apps/AppsConnect.test.tsx ui/src/pages/apps/Browse.test.tsx ui/src/lib/app-brand-assets.test.ts` plus `connector-tool-policy` and `tool-catalog-publication`: all passed after the rebase onto the split tool-access modules. - `node --test scripts/generate-unikraft-tools.test.mjs scripts/generate-daytona-tools.test.mjs scripts/app-brand-validation.test.mjs`: 20 passed. - `node scripts/check-app-brand-assets.mjs`, `node scripts/check-maps.mjs`, and `git diff --check`: passed. - `pnpm --filter @paperclipai/shared typecheck`, `pnpm --filter @paperclipai/server typecheck`, and `pnpm --filter @paperclipai/ui typecheck`: clean. - `npx playwright test --config tests/e2e/playwright.config.ts tests/e2e/unikraft-connector.spec.ts tests/e2e/daytona-connector.spec.ts`: 2 passed. - `node scripts/generate-unikraft-tools.mjs` regenerates the committed catalog with the recorded hashes. - Manual: on a local instance, open Connectors, search for Unikraft, and select **Add key**. Confirm the icon, the password field, the key guidance, and the default metro select under **Change**. - Not verified: a real Unikraft account. Live provider read and write proof is pending. ## Risks - A Unikraft token acts for the whole organization in every metro. Unikraft offers no narrower token permissions. Assign only trusted agents and review action permissions. - Sandbox actions run arbitrary commands inside the selected instance. They are classified as destructive. - A timed-out write can still complete at the provider. The bridge does not retry writes automatically. - New metros fail closed until the specification is regenerated and reviewed. Legacy `*.kraft.cloud` hosts are not reachable. - The Daytona refactor moves code without behavior changes. All Daytona tests pass without edits. - The generated JSON snapshot is large because it holds the provider input contracts. No migration or lockfile change is required. ## Model Used - Claude Opus 5.5 (Anthropic), model ID `claude-opus-5-5`, 1M context window, used through Claude Code with tool use (shell, file editing, web research, browser tests). ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge |
||
|
DottaandPaperclip
|
92cfd701c3 |
feat: add governed Braintrust app connections (#17141)
Add hosted US and EU Braintrust OAuth and API-key connections through the existing governed MCP system. Keep reviewed tool risks and catalog review policy in connector modules and a narrow registry. Preserve credential ownership, operator decisions, regional setup and changed-tool quarantine. Include official artwork, setup guidance, focused regressions and sanitized qualification evidence. US OAuth has bounded live read/write, gateway, reconnect and cleanup proof. Remaining account-method and token-lifecycle qualification gaps are documented. Co-Authored-By: Paperclip <noreply@paperclip.ing> |
