DottaandPaperclip
f69c6a8a35
## Thinking Path
> - Paperclip manages agent work and its published results.
> - GitHub review bots publish task-bound assessments through governed
tools.
> - A review can survive a caller timeout because its publication intent
is durable.
> - A transient provider failure still returned a failed receipt before
recovery.
> - Gateway replay could return an empty cached result, and publication
retries had no dedicated scheduler deadline.
> - This pull request adds bounded retries, receipt recovery, and an
independent durable worker.
> - A temporary GitHub failure can recover without another review or
duplicate publication.
## Linked Issues or Issue Description
Refs #16275. That change increased the GitHub tool deadline. This change
repairs retry and recovery behavior.
**What happened?**
A one-off GitHub HTTP 503 or 429 left the caller with a failed
publication receipt. The saved retry did not have its own scheduler
worker. An identical gateway call after a timeout could return an empty
cached result instead of the saved publication receipt.
**Expected behavior**
Temporary failures should retry under current authority. Pending work
should resume at its saved deadline after restart. Exact retries should
return the current receipt. Lost responses must not create duplicate
comments, findings, or checks.
**Steps to reproduce**
1. Admit an advisory PR review request and begin its assessment.
2. Return HTTP 503 or 429 for the first summary-comment PATCH, then let
GitHub succeed.
3. Submit the assessment through the run-scoped tool gateway.
4. On master, the result says publication failed. With this change, the
same call returns the processed receipt and neutral check.
5. For receipt recovery, use a shorter caller deadline and repeat the
exact call after publication finishes.
**Paperclip version or commit**
Reproduced on master at `6cb6ac3b8e40cb90d734ba115a2f22f850ff9477`.
## What Changed
- Classify transient transport, server, and rate-limit failures. Honor
Retry-After and primary reset deadlines. Stop retries for permanent
permission and validation errors.
- Retry safe reads and installation-token issuance within a bounded call
window. Retry publication writes through the existing outbox and
provider receipt checks.
- Wait for up to three publication attempts within a 45-second retry
window. Keep the durable eight-failure cap and enforce saved deadlines
on explicit retries.
- Register an independent GitHub publication worker with
transaction-owned wake notifications, startup recovery, and per-PR lease
deadlines.
- Recover exact comment, progress, and assessment gateway calls from
their stored publication. Require the same company, agent, task, run,
connection, tool, arguments, and current policy. Clear stale timeout
errors after recovery. Approved exact retries read the current receipt
through the existing signed replay path; the worker retains the verified
approval for deferred writes and still rejects revocation.
- Add regression coverage and update the connector guide and
feature/access maps.
## Verification
- Red: both one-off HTTP 503 and 429 regressions failed on master with a
failed publication receipt.
- Green: the same tests passed after the change.
- GitHub review workflows: 133 covered against isolated PostgreSQL 18.
The final combined run passed 132 and hit one fifteen-second timeout
under concurrent build load. All three variants of that workflow passed
in isolation. The earlier combined run passed all 131 workflows before
the two approval regressions were added.
- Retry policy and delivery coordinator tests: 35 passed.
- Additional focused checks passed for the actual eight-failure cap,
exact-argument conflicts, policy revocation, and timeout receipt
recovery. The approved deferred-publication and revocation regressions
failed before the review repair and passed after it, with one approval
request preserved.
- `pnpm -r typecheck`: passed. `pnpm build`: passed. `node
scripts/check-maps.mjs`: passed.
- The broad local `pnpm test:run` path remains blocked by the host's
exhausted PostgreSQL shared-memory slots. The changed database-backed
workflows use an isolated Docker database instead. All Paperclip CI jobs
passed at a6d81b67cb, including the required verify and E2E gates.
Greptile is 5/5 at the same head, with the approval-recovery finding
resolved. The external Superagent Security Scan also passed.
- Live acceptance passed on a dedicated single-repository local test App
and native Paperclip Runner. A test-only preload returned one HTTP 503
before the final summary PATCH. One submit_review call completed in
17.305 seconds; the bot published its 3/5 advisory review and neutral
check, then automatically finished the task. [Published
review](https://github.com/paperclipai/paperclip-permissions-smoke-20260926-pap57-fee7428e/pull/9#issuecomment-6168809007),
[neutral
check](https://github.com/paperclipai/paperclip-permissions-smoke-20260926-pap57-fee7428e/runs/115863564014).
This live test ran at de6862072c; approval-specific recovery is covered
by the later regressions. The preload is excluded from the repository
and removed from the running local instance.
## Risks
Retries make additional bounded provider requests. A provider write with
a lost response is recovered through its existing App-owned identity
before another write is permitted. The worker rechecks authority and the
PR head; revocation and superseded work stop publication. Calls can
still exceed their deadline, and repeated failures eventually exhaust
the retry cap. A pending publication is not a completed task. The agent
still owns task completion after receiving a confirmed receipt. No
schema migration is needed.
## Model Used
OpenAI Codex (GPT-6 family), with code inspection, editing, and test
execution. The exact deployment 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
- [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>
Apps, Connections, and Integrations
Audience: internal engineers and product contributors working on integrations.
Start here when adding a provider: Connection authoring runbook. It is the canonical agent tutorial from provider research and protocol classification through manifest generation, branding, secrets, deterministic tests, real-account proof, and PR submission.
Runtime authentication: AI Connections.
Long-term memory: Experimental memory connectors.
Provider notes: Google Workspace, Gmail, Asana, Sentry, PostHog, Neon, Superagent, Gauge, AgentMail, Telem.AI, iMessage Photon, and Enterpret. Optional credential custody: Vercel Connect.
Post-read action: classify a new integration request, pick the right Paperclip layer to change, and avoid creating a parallel connection framework.
Decision Record
Board decisions from PAP-13211 make the Apps v2 substrate on the PAP-10341 branch canonical:
- D1: Apps v2 is the substrate. The active model is
tool_applications,tool_connections, catalog entries, profiles, policy rules, action requests, gateway sessions, audit events, and runtime slots. Connections v1 is retired as an implementation path. - D2: one credential authority per connection, brokered projections. The
default is the Paperclip instance vault: durable third-party credentials live
in
company_secretsas secret refs. A reviewed remote MCP method may instead opt in to Vercel Connect, in which case Vercel is the durable credential authority and Paperclip stores only the connector reference and redacted grant metadata. A connection must never mix those two sources. Adapter config, plugin config, harness credential files, and run environments may receive only brokered or projected credentials. - D3: the vocabulary and three-door IA are product law. The default product doors are Apps, Connections, and Review. Protocol and operator-depth concepts live behind Developer or Advanced surfaces.
- D4: unification lands on PAP-10341. Pages, CircleBack-style harness MCP OAuth, provider gallery work, and plugin-provided integrations converge on this branch instead of spawning new integration substrates.
- D5: inbound stays thin. External clients that call Paperclip use scoped Paperclip tokens and existing profiles/rules. They do not get a separate permission model.
Canonical Object Model
Use connection as the unifying noun. A connection is four things:
- A stored credential reference.
- A capability catalog.
- A governance layer.
- An audit trail.
Everything else is an axis on that object:
| Axis | Values | It answers |
|---|---|---|
| Direction | outbound, inbound | Who is the client? |
| Transport | MCP, native REST/OpenAPI, OAuth app install, webhook | How do bytes move? |
| Auth mode | OAuth, API key/PAT, app installation, none | What does the secret represent? |
| Credential owner | company, user, run | Whose identity acts? |
| Packaging | catalog entry, plugin, skill | How does it ship? |
MCP is a transport, not a product category. "Install the Discord app", "connect Google Drive", and "add an MCP endpoint" all produce governed connections with different transport/auth values.
Layer Stack
When you are unsure where a change belongs, place it on the narrowest layer that solves the problem:
| Layer | Owns | Examples |
|---|---|---|
| Surface | user-facing Apps, Connections, Review, Developer/Advanced screens | gallery cards, setup wizard, review queue |
| Governance | profiles, bindings, allow/ask-first/block rules, quarantine, audit | read-only profile, ask-first write policy |
| Capability | action catalogs, schemas, risk classes, changed-tool review | search_issues, create_comment, schema hash |
| Credential | company_secrets, OAuth broker, credential resolver, token broker |
Slack bot token ref, Google OAuth refresh token ref |
| Identity | actor attribution and token exchange | board user, agent run, first-party service identity |
| Transport | how the external system is reached | remote HTTP MCP, local stdio, REST/OpenAPI, webhook |
The agent should not hold a durable provider credential. It should hold a Paperclip run/session token; the server or broker resolves the connection, checks governance, invokes the provider, and writes audit.
Identity vs. connections
Signing a user in and connecting a resource are different planes with different owners, different token profiles, and different homes. Do not merge them. This section is the public, connections-side statement of the identity model so connector implementers inherit the rule without depending on private identity-service documentation or re-deriving it.
| Plane | Question | Lives where | Token profile |
|---|---|---|---|
| P1. Sign-in methods | Who are you? | paperclip-id (id.paperclip.ing → Account) |
Minimal-scope provider tokens (openid email profile), used once to authenticate, encrypted at rest, never exported |
| P2. Connections (Apps) | What may your agents touch? | Paperclip App instances (tool_connections), acquired via the connect broker for hosted + self-hosted |
Rich-scope, long-lived resource tokens in the instance's encrypted vault; per-agent grants; risk-tier policy defaults |
| P3. Login with Paperclip | Who may authenticate against us? | paperclip-id OIDC provider + DB-backed client registry |
Our ES256 ID/access tokens issued by us to registered RPs (instances, the broker, future third parties) |
Everything in doc/connections/ — the First-30 matrix,
the connection authoring runbook, and the connect-broker work —
lives on plane P2. It never acquires, stores, or brokers a P1 sign-in token.
The standing rule (D7)
Adopted as a standing rule (decision D7) with the identity-model plan. State it verbatim in any P2 design so the app-store work cannot drift into merging the planes:
Sign-in tokens are never reused as resource tokens; id.paperclip.ing never stores resource tokens; no connections hub on the ID service.
P2 tokens flow broker → instance vault as pass-through only; the id.paperclip.ing Account page therefore must not grow a "Connections" hub. The reasons to hold the planes apart (from the plan §3):
- Scope discipline. Sign-in wants the narrowest grant; connections want deliberately broad ones. One button that does both is how you grant repo access just to log in.
- Blast radius. id.paperclip.ing holding every customer's Vercel/Slack/GitHub resource tokens would make it the single juiciest target in the fleet; the broker is intentionally pass-through.
- Self-hosted symmetry. Instances own their vaults, so self-hosters don't depend on our uptime to use their own connections.
- Legibility. Sign-in and connections answer different user questions, and every product we benchmarked (Vercel, Railway, GitHub, Google) keeps them on separate pages with separate names.
The explicit Vercel Connect exception does not change D7 or merge P1 and P2.
The operator chooses Vercel as the P2 credential authority for an individual
connection. id.paperclip.ing is not involved, and neither sign-in tokens nor
provider tokens pass through it. The deployment's Vercel access token or
workload OIDC identity is bootstrap authority for that external vault, not a
provider resource credential.
Naming alignment
Use the surface-correct name for each plane; they intentionally differ:
| Surface | Plane | Name to use |
|---|---|---|
| Paperclip App instances | P2 | "Connections" |
| id.paperclip.ing Account | P1 | "Ways to sign in" |
| id.paperclip.ing admin | P3 | "OIDC clients" (until the app store productizes it) |
Packaging Rule
Default to a catalog entry when an integration can be described as metadata: manifest, auth config, action catalog, resource filters, and policy defaults.
Use a plugin only when the integration needs product code such as custom UI pages, its own tables, workers, migrations, routines, or specialized ingestion. A plugin may bundle catalog entries, but it must not bypass the connection, profile, policy, credential, and audit model.
Use a skill for agent instructions. Skills may use connections; they must not own durable tokens.
Canonical Docs
- Glossary defines product and internal terms.
- Identity vs. connections is the public statement of the P1/P2/P3 boundary and the D7 standing rule for connections work.
- Security threat model harvests the keeper from PAP-2359 and maps it onto Apps v2.
- First-30 matrix harvests the keeper from PAP-2432 and is the source matrix for connector playbook work.
- Connecting any remote MCP server is the baseline: how an operator connects a standards-compliant remote MCP endpoint with no Paperclip code change, and how sign-in resolves a client.
- Connection authoring runbook is the one end-to-end, agent-executable guide for adding a vendor as a catalog entry on Apps v2: research, connection-type selection, OAuth/API-key/generated-URL setup, encrypted credential handling, branding, implementation, browser and live-provider testing, verification, and PR submission.
- Vercel Connect operator guide documents the optional external credential source, deployment flags, runtime resolution, recovery, and smoke requirements.
- MCP access governance remains the operator runbook for the current gateway, profile, policy, approval, runtime, and audit APIs.
Migration Notes
Connections v1 contributed useful policy, UX, and rollout thinking, but its
implementation branch is no longer the target. When you see old tickets or code
using connections, connection_grants, or a provider-directory mental model,
translate the intent into Apps v2:
| Connections v1 intent | Apps v2 home |
|---|---|
| Provider directory | Apps gallery / tool_applications |
| Configured provider instance | Connection / tool_connections |
| Grant allowlist | Profiles, profile bindings, policies |
| Resource filters | Policy/profile conditions plus provider config |
| Tool broker | Tool gateway and runtime supervisor |
| Connection UX tail | Apps, Connections, Review, Developer/Advanced IA |
Do not add new work to the retired v1 branch. If an old ticket still describes a valid product gap, retarget it to an active Apps v2 issue or close it as superseded with a link to the replacement.
For Slack chat bot creation, installation, recovery, and live qualification, see Automatic Slack app setup.