Repository navigation
chore: merge upstream T3 Code through 2026-08-19 - #61
Merged
Merged
Conversation
…in GitHubPullRequestCli (pingdotgg#7385) Signed-off-by: aoright <102943475+aoright@users.noreply.github.com>
Co-authored-by: shivam <91240327+shivamhwp@users.noreply.github.com>
Co-authored-by: t3-code[bot] <269035359+t3-code[bot]@users.noreply.github.com> Co-authored-by: maria <254055478+maria-rcks@users.noreply.github.com>
Co-authored-by: t3-code[bot] <269035359+t3-code[bot]@users.noreply.github.com> Co-authored-by: maria <254055478+maria-rcks@users.noreply.github.com>
…pingdotgg#7459) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…t full speed (pingdotgg#7460) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…nt (pingdotgg#6563) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…ngdotgg#7477) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
…tgg#6286) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: GPT-5.6 <noreply@openai.com>
The 2026-08-19 upstream sync added the triage command with no docs/user entry, and the docs index still called the background service Linux-only after launchd support landed. Found by PR #61 code review. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Main advanced 13 commits with thread migration and provider failover while this sync was in review. One textual conflict, in the contracts settings import list: both sides added a symbol at the same position, and both are kept. The clean merge combined incompatible logic. LimitFailoverReactor selected candidates with `config.enabled !== false`, written when every driver defaulted to enabled. This branch makes Grok, Cursor, and OpenCode default off with no explicit flag, so a never-enabled instance would have passed that check and received a failing-over thread on a provider the server never probed or registered. Candidate selection now goes through resolveProviderInstanceEnabled, extracted as a pure failoverCandidates alongside the existing remainingScore and covered by a regression test that fails against the old predicate. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
roughcoder
added a commit
that referenced
this pull request
Aug 20, 2026
PR #61 was squash-merged, so its merge commit's second parent was dropped and upstream's history stopped being reachable from main. That left `git merge-base main upstream/main` reporting a4cc136, the base of the 2026-08-17 sync, so the next integration would have tried to replay the 28 commits #61 already integrated and re-litigated every Phoenix adaptation in them. This merge carries no file changes; the tree is exactly 7e40aff. Its only effect is to make f2d5fc9 an ancestor of main again so merge-base resolves correctly for future syncs. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This was referenced Aug 20, 2026
roughcoder
added a commit
that referenced
this pull request
Aug 20, 2026
The suite runs ~9m30s against a 10m cap, so ordinary runner variance cancelled healthy runs. PR #61 lost one at 10m02s with every test already passed, killed during post-job cleanup, and the identical rerun passed at 9m21s. Raises the test job to 12 minutes. The other jobs keep their 10m cap: Check runs ~3m and Release Smoke ~1m, and a cap they never approach is still doing its job of catching a hang. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
roughcoder
added a commit
that referenced
this pull request
Aug 20, 2026
Squash-merging an integration PR keeps the files and drops the upstream parent, so the next sync's merge base silently rewinds to the sync before last and replays commits that were already integrated. #61 hit this; the ancestry was restored afterwards in 3584603. Adds the recovery step to the runbook, and a check in the pin step so a rewound merge base is caught before any conflict is resolved rather than after. Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
roughcoder
added a commit
that referenced
this pull request
Aug 20, 2026
The upstream merge in #61 moved @clerk/electron-passkeys onto a canary release whose platform packages ship without their napi binaries: every 0.0.4 canary tarball is a 1.6 KB stub of package.json + LICENSE, yet package.json still declares the absent .node file as "main". Staging the mac app then fails to resolve @clerk/electron-passkeys-darwin-arm64, so the Phoenix Build workflow has been red on every push since. Pin @clerk/electron to stable 0.0.34 (the same base version as the canary) and @clerk/electron-passkeys to 0.0.3, whose platform packages carry real binaries. Stable 0.0.34 peers on electron-passkeys "*" and so accepts 0.0.3, where the canary pinned the exact canary build. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
15 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Twice-weekly upstream sync. Phoenix
ba2e6449was 28 commits behindpingdotgg/t3code(f2d5fc91, basea4cc1367), and we want every compatible upstream improvement without giving up Phoenix identity or the distribution quarantine.Resolved as a three-way semantic merge per
docs/operations/upstream-integration.md: 11 textual conflicts resolved, 34 both-branches-changed files audited. Headline decisions: adopted the launchd background-service backend with Phoenix runtime identity (com.goodbird.phoenix.service,PHOENIX_HOME) while keeping install/update/onboarding disabled; adapted the newtriageCLI to route issues and source lookups togoodbirdhq/phoenixinstead of upstream; merged Phoenix's Sessions surface into upstream's new data-driven shortcut menu. Fixed two clean-merge defects (launchd test fixtures and a GitManager test asserting pre-Phoenix identifiers/field lists). No release automation re-enabled.Full record:
docs/operations/upstream-integrations/2026-08-19.md. Verified with per-package typechecks (server, web, desktop, contracts, shared, client-runtime) and ~670 focused tests across every conflict and overlap area.Produced by Claude Fable 5 via Claude Code.
🤖 Generated with Claude Code
Need help on this PR? Tag
@codesmith-botwith what you need. Autofix is disabled.