Files
PaperClipAI/server
DottaandPaperclip ba00e9ef1a refactor(server): isolate provider webhook handling (#16183)
## Thinking Path

> - Paperclip connects provider messages to governed agent work.
> - Webhooks must authenticate and persist messages before acknowledging
them.
> - The shared service still contains GitHub and Slack protocol details.
> - The platform-boundary rule places those details in provider folders.
> - This pull request extracts provider webhook handling behind typed
methods.
> - The service retains dispatch, runtime selection, and durable
lifecycle recording.

## Linked Issues or Issue Description

Depends on #16180; review this PR against its parent branch. Refs
#16080. Related open work: #16163 changes GitHub automatic-author
policy. This extraction preserves that policy implementation.

**What existing behavior does this improve?**
The ownership and readability of GitHub and Slack webhook handling.

**Current behavior**
The shared service implements GitHub signatures, replay fences,
automatic event messages, and Slack callback verification.

**Proposed behavior**
Place those details in the respective provider folders and keep the
common dispatcher small.

**Reason and benefit**
Provider code can be reviewed without reading the shared task service.

## What Changed

- Extract GitHub request-local deadline and replay scope (60 lines).
- Extract GitHub authentication and durable-ingress coordination (280
lines).
- Extract GitHub event admission, message construction, and policy-skip
receipts (140 lines).
- Extract Slack workspace validation and authenticated callback
observation (131 lines).
- Keep the same service-owned background task set and GitHub provenance
maps. Forward late-initialized callbacks without reading them during
construction.
- Reduce chat-channels.ts from 7,383 to 6,880 lines: 503 net lines
removed.

## Verification

- Merge preparation: replayed this stack onto current master after
#16180 landed. Preserve master's new Telegram scheduling methods and
connection-scoped Photon notifications. Server typecheck, callback
syntax audit, and module-boundary check passed. Fresh integration, CI,
and current-head review are running. Earlier verification below
describes the previous stack commits.

- Parent baseline: 544 selected GitHub, Slack, and webhook integration
cases passed; 751 other cases excluded.
- Initial extracted run: 543 passed and one idle-queue scan assertion
failed. The idle-queue case passed separately on both the parent and
extracted code. Its publication implementation is unchanged. The full
selected-suite retry passed all 544 cases; the initial failure remains
documented and its cause is not established.
- Server typecheck, module-boundary check, and git diff --check passed.
- Syntax-tree comparison verifies unchanged request-scope logic, durable
ingress, leased authentication, event handling, workspace validation,
and callback recording. All unrelated service functions remain
unchanged.
- Repository typecheck and production build passed for the parent
extraction. Full-stack verification will run at the final handoff. The
first broad local run hit a macOS PostgreSQL shared-memory fixture limit
and was interrupted after that failure.
- All 55 current-head CI checks passed. Greptile: 5/5 on f74ecb5094,
with zero inline findings. No live third-party execution is claimed.

- Previous stack validation on #16192 (55417debfe): repository typecheck
and build passed; all 1,295 chat integration cases plus 14 source-wiring
checks passed locally (1,309 total). All five PRs have current-head 5/5
reviews and 55 passing CI checks. The broad local pnpm test:run hit
PostgreSQL shared-memory fixture failures and was interrupted after 868
seconds (exit 130); it is not a local full-suite pass.

## Risks

Moving callback code can change initialization timing or lose
request-scoped state. The new factories retain live option getters, the
exact runtime context and provenance maps, and typed forwarding for late
initialization. Authentication stays inside the existing credential
lease. The common lifecycle write still completes before returning
provider success.

## Model Used

OpenAI GPT-6 via Codex, with code execution and repository tools. The
precise serving variant 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
- [ ] 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>
2026-10-13 11:02:19 -05:00
..