Co-Authored-By: Paperclip <noreply@paperclip.ing>
* codex/claude-runner-default:
perf(chat): stop polling failed run retries (#15871)
docs(release): stable notes for the 2026.1012.1-beta.0 soak (v2026.1015.0) (#16053)
Support test Cloud connector identities (#16033)
fix(inbox): keep automatic recovery cleanup out of personal participation (#16063)
Co-Authored-By: Paperclip <noreply@paperclip.ing>
* commit 'dae808fe02694d13d322c535944910813b21f6d6':
perf(chat): stop polling failed run retries (#15871)
docs(release): stable notes for the 2026.1012.1-beta.0 soak (v2026.1015.0) (#16053)
Support test Cloud connector identities (#16033)
fix(inbox): keep automatic recovery cleanup out of personal participation (#16063)
Merge current master and reuse the durable-resume test envelope correction. Wait for actual callback entry under the existing recovery deadline and preserve original early errors.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Retain both Claude SDK admission and bounded Intel initial-install failure evidence. All25 focused consumer controls pass; no provider, build campaign, local Rust or Docker.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Preserve Claude v7 effort, lease ownership, historical readers, and Pi25 source binding while composing canonical16a and task-workspace masterf329. Protocol/surfaces noemit and all generation checks pass.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Stop one-second scans of failed chat-run retries. Wake the existing retry queue after transaction commits and at saved retry or stale-claim deadlines. Preserve claim fences, authorization, recovery, and idempotent agent admission.
Resolve conflicts with the current chat-service extraction by applying the retry signals and shared deadline predicate in the failed-run retry module.
Verified with pnpm typecheck, pnpm build, all 1,328 chat and coordinator regression tests, and current-head CI. Greptile reviewed the final head at 5/5 with no unresolved threads.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Preserve guarded test-home cleanup and both Storybook source reviews. Keep the installer deadline and all artifact/lifecycle/runtime assertions intact while retaining its failed npm progress.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work
> - The release process promotes a soaked beta to stable. The stable
GitHub Release body comes from `releases/beta/v<beta-version>.md` on
`master`
> - The `draft_stable_notes` job pushed a commit-log skeleton to this
branch when `2026.1012.1-beta.0` published today (workflow run
38311557959)
> - The skeleton is grouped commit subjects. It is not in release-notes
voice and it does not tell a self-hoster what to do before an upgrade
> - This pull request replaces the skeleton with curated notes. Each
claim traces to a pull request in `v2026.1012.0..1b4c8038`
> - The benefit is that the earliest stable promotion (v2026.1015.0)
finds finished notes on `master` on time
## Linked Issues or Issue Description
**Issue type**
Missing content
**Where is the issue?**
`releases/beta/v2026.1012.1-beta.0.md` — the stable-notes file for the
`2026.1012.1-beta.0` soak. It exists only on this machine-owned branch
and only as the auto-generated skeleton.
**What's wrong?**
The stable promotion reads this file from `master` and publishes it as
the GitHub Release body. Until this branch merges, the stable preflight
has no notes to resolve. The skeleton lists raw commit subjects with
nested PR summaries. It does not call out the three changes that need
operator attention before the upgrade.
**Suggested fix**
Merge the curated notes so the notes invariant holds for the
v2026.1015.0 promotion. Correct the `> Released:` date in a follow-up if
the promotion date slips.
## What Changed
- Replaced the skeleton in `releases/beta/v2026.1012.1-beta.0.md` with
curated stable notes for v2026.1015.0 in the same layout as
`releases/beta/v2026.1006.0-beta.0.md`: overview, Breaking Changes,
Highlights, Fixes, Improvements, Upgrade Guide, Contributors
- Breaking Changes lists the expanded Operator role preset (#15377),
GitHub review bots behind a new default-off experimental setting
(#15416), and budget policies that hold on unpriced usage by default
(#14997)
- Upgrade Guide enumerates migrations `0306` through `0333` with
one-phrase effects, the three new environment variables with defaults,
the three new default-off experimental settings, the new agent identity
key variables for custom adapters, and the private-task API behavior
- Release and CI internals, smoke specs, canary tooling, Paperclip Cloud
work, heartbeat refactors, dependency bumps, and test-only changes are
left out. Features already described in the v2026.1012.0 notes are not
repeated
## Verification
- Beta publish is complete: npm dist-tag `beta` is `2026.1012.1-beta.0`,
tag `beta/v2026.1012.1-beta.0` points at
`1b4c803816e82b0265c8422e1a804327a7ed8c54`, and the `publish_beta`,
`smoke_beta`, and `draft_stable_notes` jobs in run 38311557959 all
succeeded
- `git rev-list --count
v2026.1012.0..1b4c803816e82b0265c8422e1a804327a7ed8c54` returns 200,
which matches the overview and Contributors section
- `git shortlog -sn --no-merges` over the same range shows 6 human
authors after the two bot accounts are excluded. The two external
handles were read from the GitHub commit authors, not guessed from names
- `./scripts/release.sh stable --date 2026-10-15 --print-version`
returns `2026.1015.0`
- Every `#NNNN` link in the file resolves to a pull request inside the
range. Migration numbers, setting keys, env var names, and defaults were
checked in the code on the source commit, not in commit subjects
- The file contains no internal ticket ids or instance-local links
- The compare view for this branch shows one added file and no source
changes
## Risks
- Low risk: a single markdown file, no source changes. If the promotion
date slips past 2026-10-15, the `> Released:` line and H1 need a
one-line update before the stable dispatch. The beta-keyed filename
makes that re-date harmless
## Model Used
- Claude (Anthropic), model ID `claude-fable-5-1` (Claude Fable 5.1),
extended thinking enabled, tool use via Claude Code
## 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
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: devinfoley <139239+devinfoley@users.noreply.github.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work.
> - Its Cloud connector uses an environment-bound identity to call a
hosted broker.
> - The identity parser and enrollment flow accept development, staging,
and production.
> - A separate test broker needs its own environment identity.
> - This pull request accepts test identities and maps the test broker
to that environment.
> - Known hosted brokers still reject identities from other
environments.
## Linked Issues or Issue Description
**What existing behavior does this improve?**
Cloud connector configuration and self-host enrollment.
**Current behavior**
The app rejects an identity with environment `test`. Self-host
enrollment infers development for the test broker.
**Proposed behavior**
Accept `test` identities at `https://test.paperclip.app`. Infer test
during self-host enrollment. Reject development, staging, and production
identities at that broker. Continue to reject test identities at the
production and staging brokers.
**Reason and benefit**
Branch testing needs a separate broker so test identities and requests
stay in their selected environment. The existing identity, enrollment,
and signature flow can provide that isolation with a new allowed
environment value.
**Breaking changes**
Existing environment values and broker checks keep their current
behavior. Older app versions cannot use test connector identities.
## What Changed
- Add `test` to managed and self-hosted connector environment types and
validation, and to the board API enrollment status type.
- Infer test from its hosted broker and enforce the matching
environment.
- Test enrollment persistence, signed request environment/audience
claims, and broker mismatch rejection.
- Document test broker configuration in the development guide.
## Verification
- Connector, enrollment, and board API-client suites passed: 63 tests
across three files. Board UI typecheck and token gates passed after
synchronizing its enrollment response type.
- `pnpm -r typecheck` and `pnpm build` passed, including after rebasing
onto current master.
- `pnpm test:run` completed its general server stage: 17,867 passed, 105
skipped, and 9 failed across 5 files. The runner stops at that failing
stage, so it did not run subsequent workspace/serialized stages locally.
All nine failures also reproduce on the unchanged base commit in a
separate checkout. Five skill-related cases resolve an unrelated
ancestor skills directory; two workspace cases accept a remote main
branch that their fixture expects to be absent; two permission-cleanup
cases return macOS `ENOTEMPTY` instead of the expected permission error.
- Current-head GitHub CI passed, including full general and serialized
test suites, typecheck, build, runner checks, and browser e2e. Greptile
is 5/5 with no unresolved comments.
- `git diff --check` and Gitleaks passed.
## Risks
The new environment value is additive. The test broker must support test
identities before use. This change preserves existing signature
handling, managed identity overrides, and rejection of mismatched
identities at known hosted brokers. A local full-suite pass is not
claimed; the nine failures above also occur without this change.
## Model Used
OpenAI Codex, GPT-6, with reasoning and tool 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
- [ ] 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>
Co-authored-by: Dotta <34892728+cryppadotta@users.noreply.github.com>
## Thinking Path
> - Paperclip is the open source app people use to manage AI agents for
work.
> - The personal inbox uses issue activity to identify a person's
participation.
> - Issue reads can cancel stale recovery actions and record that
cleanup as the reader's action.
> - This adds unrelated tasks to the inbox and refreshes their activity
timestamps.
> - A workspace owner hold can also alternate between creation and
cancellation after a task finishes.
> - This pull request corrects cleanup attribution, excludes historical
read cleanup, and retains one exact owner hold.
> - Agents can read authorized issue activity to diagnose the same
problem themselves.
## Linked Issues or Issue Description
Refs #15742 excludes logs by the execution-recovery actor from activity
timestamps. This PR covers reader-attributed stale recovery cleanup,
personal inbox participation, exact workspace owner holds, and sandbox
issue activity access. I searched open issues and PRs for recovery inbox
and activity bridge fixes. I found no duplicate for this defect.
**What happened?**
Loading issue lists cancels stale recovery actions under the reader's
identity. The personal inbox treats those rows as user participation.
One completed issue had hundreds of such rows and no edits or comments
by the affected reader. A retained native workspace owner repeatedly
creates a verification action that read projections then cancel. Sandbox
agents cannot inspect the issue activity because the reverse bridge
rejects the route.
**Expected behavior**
Automatic cleanup belongs to the system and does not add an issue to a
reader's inbox. Historical cleanup rows remain auditable but do not
count as participation or recent activity. An unresolved physical
workspace owner keeps one stable verification action. Agents can inspect
activity only for issues their identity may read.
**Steps to reproduce**
1. Finish a task whose native run retains an unverifiable workspace
owner receipt.
2. Repeat workspace recovery checks and GET requests for issue detail,
compact lists, and recovery actions.
3. Observe new recovery action IDs and reader-attributed cancellations
in the activity history.
4. Query the personal inbox for that reader and observe the unrelated
task.
5. From a sandbox agent, request GET /api/issues/:id/activity and
observe Route not allowed.
**Paperclip version or commit**
Observed at f5346c05b9. Regression tests
use current master behavior.
**Deployment mode**
Authenticated hosted instance with native and legacy sandbox agents. The
inbox defect is independent of the adapter.
## What Changed
- Attribute automatic stale recovery cleanup to the recovery system
actor.
- Exclude exact historical read cleanup rows from personal inbox
membership, unread counts, and both activity timestamp queries. Keep all
audit rows and explicit user recovery decisions.
- Retain owner verification holds only while the exact company, run,
issue, and owner fingerprint still match. Apply the same rule to reads
and background recovery.
- Make repeated checks for the same owner hold leave its identity,
attempts, and timestamps unchanged. Keep physical takeover blocked.
- Allow GET issue activity through the sandbox bridge. Keep the existing
company and issue-read authorization and reject broader audit routes.
- Add regressions and update the runtime API and recovery documentation.
Refresh generated source references for the documentation edits. Update
the paired activity, inbox, and recovery maps.
## Verification
- Focused Vitest verification: 315 tests pass across five files. The
updated terminal owner hold regression also passes. The files cover
issue service, recovery routes, activity authorization, sandbox callback
bridge, and native workspace finalization.
- `pnpm -r typecheck` and `pnpm build` pass locally. The full general
and serialized unit test matrix passes in CI on the latest commit. The
duplicate local `pnpm test:run` was stopped after about 30 minutes once
that full CI coverage passed; it did not report a local full-suite pass.
- `node scripts/check-maps.mjs` passes for all paired areas and
capability IDs.
- Apex is 5/5 on commit 44c3787888. Its
access map finding is fixed and resolved. All CI checks are green on
this commit.
- The new regression repeats reads and background sweeps for done and
cancelled tasks. It checks one unchanged hold, zero physical copyback,
zero personal inbox participation, and system cleanup after the exact
receipt is cleared.
- Historical cleanup tests preserve explicit recovery decisions,
nullable metadata, unread counts, and the original audit rows.
- Authorization tests permit an authorized agent and reject
cross-company or issue-boundary reads. Bridge tests reject write
methods, extra path segments, and company-wide audit routes.
## Risks
Existing inbox items can disappear if their only participation evidence
was automatic read cleanup. No data migration or audit deletion is
needed. An unresolved workspace owner remains visible after task
completion; this preserves the physical writer fence. A stale or cleared
owner does not retain the hold. Issue activity access expands only the
sandbox transport allowlist and still uses existing controller
authorization.
## Model Used
OpenAI GPT-6 through Codex, with reasoning, repository inspection, code
editing, shell execution, and test tools. The session identifies the
model family as GPT-6. A more specific runtime model ID and context
window size are 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>
Co-authored-by: Amy (Opus) <noreply@anthropic.com>
Keep the qualified four-target selection and both Linux provider-pack assembly paths while carrying the current shared source.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Preserve all existing story IDs, classify scoped source review separately from six fixture plays, and retain final package/provider qualification gates.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Give tasks durable company-scoped workspace bindings independent of project assignment, channel defaults and authorized repository preparation. Reuse verified input seeds and sparse output recovery while preserving trust, privacy and controller provisioning boundaries.
Full CI and E2E, Superagent, legacy checks and Greptile Apex 5/5 passed on reviewed head 753680b61c. All review findings are resolved; the PR records focused verification and historical acceptance limits.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Preserve canonical startup ownership, scoped evidence, current Storybook IDs, and the Codex form controls while carrying Claude effort and immutable source bindings.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Extract the guided setup test fixture without dropping cases. Supply current settings/discovery contracts and Opus 5.5 in the isolated Storybook boundary.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Retain Pi profile 24 and bind the updated sandbox source in profile 25. Provider closures and qualification boundaries are unchanged.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Co-Authored-By: Paperclip <noreply@paperclip.ing>
* codex/claude-runner-default:
docs: record retained-preview mobile runner evidence
fix(runner): retain startup ownership through cancellation
test(server): initialize issue routes within the fixture hook
test(runner): reserve the occupied credential quorum fixture
fix(ui): wait for current guided runner settings
Preserve the source prerequisite repair and Claude effort/resume contracts. Eight no-process controls and runner noemit pass; native readiness remains a hosted CI gate.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
Keep feature and access maps synchronized with the scoped cancellation repair, and retain the exact source and unresolved installed/staging qualification.
Co-Authored-By: Paperclip <noreply@paperclip.ing>
