Skip to content

chore: sync upstream (2026-11-02) - #4

Merged
getBoolean merged 24 commits into
mainfrom
sync/upstream-2026-11-02
Nov 2, 2026
Merged

getBoolean merged 24 commits into
mainfrom
sync/upstream-2026-11-02

Conversation

@getBoolean

Copy link
Copy Markdown
Owner

Merges upstream pingdotgg/t3code main into the fork's main: 267c90831a..47f2c8d29e from upstream's side (23 upstream commits, through pingdotgg#29169).

Conflicts

None. The merge was clean.

Fork adaptations

  • Runner audit (fork_ci_audit.py --rewrite-runners --prefix T3CODE) found no new hard-coded runner labels.
  • No new github.repository == 'pingdotgg/t3code' guards. The count is unchanged at 17.
  • Upstream ci(release): preview releases run from any branch pingdotgg/t3code#27828 changed release.yml so preview releases run from any branch (the environment is now empty for preview, and Clerk/relay ids come from .env.example). This doesn't affect the fork's nightly path. The fork's T3CODE_RELEASES gate, UPSTREAM tracing skip and installer repository rewrite are all still there.
  • The merge didn't touch any per-project git account files (gitAccounts, ProviderSessionManager, ProviderTurnStartService, GitManager, terminal Manager, PullRequestService, MCP project toolkit). It also didn't touch the release-repository files (cliRelease.ts, public-config.ts, npmRelease.ts).

Verification (Windows)

  • vp run check:changed: format, lint, knip, graphql, typecheck (39 packages) and react compiler all pass. The test step hit the local 900s timeout and is being re-run with a longer one. CI runs the full suites.
  • Fork suites, apps/server (src/gitAccounts, terminal Manager, cloud npmRelease, PullRequestService, src/mcp/toolkits/project): 9 files, 255/255 passed.
  • Fork suite, apps/web GitAccountsSettings.browser.test.tsx: 2/2 passed.

Model: Claude Opus 5.5 (claude-opus-5-5), harness: Claude Code in T3 Code.

Yash-Singh1 and others added 24 commits November 1, 2026 21:37
…r Escape (pingdotgg#28298)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: maria-rcks <maria@kuuro.net>
Co-authored-by: shivam <91240327+shivamhwp@users.noreply.github.com>
Co-authored-by: PR Batch Tester <agent@local.test>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…e Mode (pingdotgg#29026)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ct (pingdotgg#29068)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…is slow on a loaded host (pingdotgg#28201)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… reseller marks them up (pingdotgg#28004)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…er update (pingdotgg#27992)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…28995)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
… work (pingdotgg#28629)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…29053)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…server (pingdotgg#29127)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Completion alerts waited two seconds after the client received a completed thread state. Audio was also fetched and decoded only when an alert needed to play, adding a first-play delay.

Notify immediately when the thread shell reports completion, and preload both sounds when a user gesture unlocks audio. Keep completion deduplication and suppression for existing, archived, and stopped runs.

This intentionally removes the continuation grace for completion alerts. A completed run can now notify before an automatic continuation starts. The sidebar's two-second placement grace and server settlement behavior remain in place.

Validation: `vp run check:changed` passes after merging current main, including 40 affected tests, formatting, lint, unused-code checks, affected typechecks, and React compiler checks. Tests cover immediate completion, duplicate suppression across reconnects, preloaded audio reuse, and retrying failed loads on alerts without retrying on every gesture. Audible timing has not been verified in a live client.

Model: GPT 6.1 Sol. Harness: Codex through T3 Code.
…unched T3 (pingdotgg#29169)

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions github-actions Bot added vouch:trusted PR author is trusted by repo permissions or the VOUCHED list. size:XXL labels Nov 2, 2026
@github-actions

github-actions Bot commented Nov 2, 2026

Copy link
Copy Markdown

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider Metric Main baseline This PR Impact PR ceiling
Codex Total thread wire 5.0 KiB 5.0 KiB 0 B (0.0%) 6.8 KiB ✅
Codex Thread snapshot wire 3.8 KiB 3.8 KiB 0 B (0.0%) 4.9 KiB ✅
Codex Live turn WebSocket wire 1.2 KiB 1.2 KiB 0 B (0.0%) 2.0 KiB ✅
Codex Live turn WebSocket decoded 21.6 KiB 21.6 KiB 0 B (0.0%) 24.4 KiB ✅
Codex Live turn messages 2 2 0 (0.0%) 8 ✅
Claude Total thread wire 5.0 KiB 5.1 KiB +6 B (+0.1%) 6.8 KiB ✅
Claude Thread snapshot wire 3.8 KiB 3.8 KiB 0 B (0.0%) 4.9 KiB ✅
Claude Live turn WebSocket wire 1.3 KiB 1.3 KiB +6 B (+0.5%) 2.0 KiB ✅
Claude Live turn WebSocket decoded 21.9 KiB 21.9 KiB 0 B (0.0%) 24.4 KiB ✅
Claude Live turn messages 2 2 0 (0.0%) 8 ✅

Baseline: 267c908 · PR result: df3db0e · 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: 108.5 KiB
  • Claude decoded thread snapshot: 108.8 KiB

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

@getBoolean
getBoolean merged commit aa51678 into main Nov 2, 2026
25 checks passed
@getBoolean
getBoolean deleted the sync/upstream-2026-11-02 branch November 2, 2026 12:31
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

size:XXL vouch:trusted PR author is trusted by repo permissions or the VOUCHED list.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

7 participants