perf(server): pull request sync reads only threads with linked pull requests

[t3dotgg]

PullRequestSyncReactor read the full shell snapshot on every 1-minute sweep and on every new PR link. On a server with about 5,000 threads and 130 projects, one full read runs 6 queries (most of them over every active thread), resolves repository identity for every project (git spawns on cache misses), and decodes the whole snapshot again. The reactor only uses the threads that have a linked pull request, and only 5 fields of each.

Fix

• New ProjectionSnapshotQuery.listThreadsWithPullRequests(). It is one SQL query that joins projection_thread_pull_requests to active (not deleted, not archived) threads. This is the same thread filter as the shell snapshot.
  • It returns id, projectId, settledOverride, settledAt and pullRequests. These are all the fields the reactor reads.
  • It keeps the shell snapshot's thread order and link order. So PR groups, and the project used for each host read, do not change.
  • It does not resolve repository identity. Identity only feeds the legacy linkedPullRequest field, and the reactor does not read it.
• The reactor reads this query instead of getShellSnapshot(). Its archivedAt check is gone because the query already skips archived threads.
• The other test files only add the new method to their full ProjectionSnapshotQuery mocks.

Verification

• New SQL test in ProjectionSnapshotQuery.test.ts. It seeds active, settled, unlinked, archived and deleted threads across 2 projects. It checks that the new query gives the same threads, fields and order as the shell snapshot's linked threads, in 1 SQL statement, with 0 identity lookups. It fails if the sort order changes or if the archived filter is removed.
• PullRequestSyncReactor.test.ts: the sync still writes the snapshot to a linked PR thread, and getShellSnapshot runs 0 times. It fails against the old reactor code.
• vp test run on all touched test files: 284 of 285 pass. The 1 failure (AgentSessionScanner symlink test) also fails with main's copy of that file on this Mac.
• vp lint and vp fmt on the changed files, and vp run --filter t3 typecheck.

Related: #13691 (the same kind of fix for settlement and PR discovery).

Made by Claude Opus 5.5 (1M context) in Claude Code, running in T3 Code.

🤖 Generated with Claude Code

Summary by CodeRabbit

• Improvements
  • Pull-request synchronization now processes active threads with linked pull requests while preserving their established order and the thread, project, settlement, and pull-request details needed for processing. Syncing these links no longer requires loading a full project snapshot, while existing synchronization behavior remains unchanged.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.5 KiB	13.5 KiB	+12 B (+0.1%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.1 KiB	7.1 KiB	−5 B (−0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.5 KiB	6.5 KiB	+17 B (+0.3%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	56.3 KiB	56.3 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	10	10	0 (0.0%)	21	✅
Claude	Total thread wire	13.5 KiB	13.5 KiB	−15 B (−0.1%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.1 KiB	7.1 KiB	−4 B (−0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.4 KiB	6.4 KiB	−11 B (−0.2%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.0 KiB	57.0 KiB	0 B (0.0%)	66.4 KiB	✅
Claude	Live turn messages	9	9	0 (0.0%)	21	✅

Baseline: 2a9832b · PR result: e2cb766 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 114.0 KiB
• Claude decoded thread snapshot: 114.7 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Approved at e2cb766

Macroscope's review found this PR approvable — This is a contained server-side performance optimization that replaces full shell hydration with a direct query for active threads that already have pull-request links. Existing synchronization and persistence behavior is preserved, with focused tests covering ordering, filtering, and the reduced query path.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: pingdotgg/t3code/.coderabbit.yaml

Review profile: CHILL

Plan: Team

Run ID: 1dbc9645-9799-45d0-8c0c-710209960672

📥 Commits

Reviewing files that changed from the base of the PR and between b8cd221 and e2cb766.

📒 Files selected for processing (1)

• apps/server/src/orchestration/Layers/ProjectionSnapshotQuery.ts

Limit details: You’ve used all 10 included reviews currently available.

---

📝 Walkthrough

Walkthrough

The projection snapshot query now returns active threads with pull-request links. The pull-request sync reactor uses this query instead of reading a full shell snapshot. Tests cover the query results and verify that the sync sweep does not read a full shell snapshot.

Changes

Pull-request projection and sync

Layer / File(s)	Summary
Define and implement the pull-request thread query apps/server/src/orchestration/Services/ProjectionSnapshotQuery.ts, apps/server/src/orchestration/Layers/ProjectionSnapshotQuery.ts, apps/server/src/orchestration/Layers/ProjectionSnapshotQuery.test.ts, apps/server/src/checkpointing/CheckpointDiffQuery.test.ts, apps/server/src/orchestration/Layers/OrchestrationEngine.test.ts, apps/server/src/project/*, apps/server/src/provider/Layers/*, apps/server/src/serverRuntimeStartup.test.ts	Adds the query contract and implementation. It returns ordered records for active threads with pull-request links and maps query failures to persistence errors. Tests cover the results. Affected query test doubles add unused-operation stubs.
Use the projection query in pull-request sync apps/server/src/orchestration/PullRequestSyncReactor.ts, apps/server/src/orchestration/PullRequestSyncReactor.test.ts	The sync sweep reads projection thread records instead of a full shell snapshot. The test asserts that the sweep does not read a full shell snapshot.

Priority: ➖ Normal

Estimated code review effort: 2 (Simple) | ~15 minutes

Change: Refactor

Sequence Diagram(s)

sequenceDiagram
  participant PullRequestSyncReactor
  participant ProjectionSnapshotQuery
  participant Database
  PullRequestSyncReactor->>ProjectionSnapshotQuery: listThreadsWithPullRequests()
  ProjectionSnapshotQuery->>Database: select active threads and pull-request links
  Database-->>ProjectionSnapshotQuery: ordered query rows
  ProjectionSnapshotQuery-->>PullRequestSyncReactor: grouped thread records
  PullRequestSyncReactor->>PullRequestSyncReactor: group visible pull-request links

Loading

Suggested reviewers: juliusmarminge

Merge Risk: ⚪ Minimal · up to e2cb7

The sync sweep retains the verified thread eligibility behavior. No issue requiring a fix before merge was identified.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 3 functions across 12 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title clearly and concisely describes the main performance change: pull request synchronization now reads only threads with linked pull requests.
Description check	✅ Passed	The description explains the problem, implementation, scope, and verification results. It omits the template headings and checklist, but it provides the required substantive information and is mostly …

• Fix all pre-merge checks with AI

✨ Finishing Touches 💡 1

📝 Generate docstrings 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

_{Comment @coderabbitai help to get the list of available commands.}

Dismissing prior approval to re-evaluate e2cb766