Repository navigation
chore(fork): merge upstream 701e13ceb into atli - #49
Merged
Merged
Conversation
Co-authored-by: maria-rcks <maria@kuuro.net>
…he server attached (pingdotgg#27456) Co-authored-by: maria-rcks <maria@kuuro.net>
…ingdotgg#27472) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ingdotgg#27485) Co-authored-by: maria-rcks <maria@kuuro.net>
…gg#27497) Co-authored-by: maria-rcks <maria@kuuro.net>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: maria-rcks <maria@kuuro.net>
Co-authored-by: maria-rcks <maria@kuuro.net>
…7635) Co-authored-by: maria-rcks <maria@kuuro.net>
…pingdotgg#27662) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…rly 404 (pingdotgg#27722) Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…7757) Co-authored-by: maria-rcks <maria@kuuro.net>
…branch named t3 exists (pingdotgg#27815) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…otgg#27848) Co-authored-by: maria-rcks <maria@kuuro.net>
…n alias (pingdotgg#27859) Co-authored-by: maria-rcks <maria@kuuro.net>
Co-authored-by: Omar <76748259+omar9848492@users.noreply.github.com>
…s hanging (pingdotgg#27924) Co-authored-by: Trevor Edwards <9443483+trevoedwards@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ingdotgg#27990) Co-authored-by: Claude Code (claude-opus-5-5) <claude-code@logueos.local> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…T3 runs (pingdotgg#28064) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…op (pingdotgg#28072) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…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>
…ngdotgg#28983) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ngdotgg#28520) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ngdotgg#28387) Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…pingdotgg#28314) 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>
pingdotgg#28102) 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>
…ingdotgg#29017) 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>
Merges 745 upstream commits from pingdotgg/t3code main (32ecd71..701e13c). Conflict resolutions: - Claude MCP elicitation: upstream added its own approval-card handler; keep the fork's form-capable handler and drop upstream's duplicate, adopting its broader one-time choice spellings and no-feedback card for elicitations. - Runtime instructions: keep the fork's watch-off variant alongside upstream's session-invariant PULL_REQUEST_LINKING_INSTRUCTIONS and pullRequestLinking. - OpenCode: carry the fork's pull request watch flag on the turn now that upstream builds the system prompt per prompt. - Draft run-stop loading: take upstream's threadDetailRef fix over the fork's. - Self-update: keep the fork's staged preflight timeout; add upstream's managed update policy and launcher recording. - check:changed: keep NUL-safe gitPaths; take upstream's merge-base and test split. - Background work label: upstream "Waiting · <label>" with the fork's always-waiting state. - Update fork tests for upstream's ManagedPolicy, BootService.stop and repointLauncher(baseDir) changes; refresh agentContextBudget.json. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Skald review statusSkald: review ready (success) at review state version 11, head BlockingNothing is holding this pull request open. ClearedNothing is cleared at this epoch. Blocking: 0. Advisory: 0. Cleared: 0. Fenced: 0. Review rounds
|
There was a problem hiding this comment.
Skald review is clean
Reviewed head 92b68f74ca2f29b60c99a5fcfcffccd02b0a127c against base atli (8a850f51bcb8e4c5c005b1053407869ebfaf3071).
First round
Standing summary: open
Review details
Completed reviewers:
- architecture (hub-dispatch-v2:claude:claude-sonnet-5-5:medium:8ef92ce46d3e55c4)
- frontend (hub-dispatch-v2:claude:claude-sonnet-5-5:medium:8ef92ce46d3e55c4)
- security (hub-dispatch-v2:claude:claude-sonnet-5-5:medium:8ef92ce46d3e55c4)
- testing (hub-dispatch-v2:claude:claude-sonnet-5-5:medium:8ef92ce46d3e55c4)
This was referenced Nov 2, 2026
gabynevada
added a commit
that referenced
this pull request
Nov 2, 2026
chore(fork): restore upstream ancestry lost in the #49 squash merge
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
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.
Brings
atliup to date with upstreampingdotgg/t3codemain: 745 commits,32ecd71726..701e13ceb.Notable upstream changes
t3 updateand the self-updater.link_pull_requests/ bulk unlink;preview_opengoes straight to a dev-server port; Claude MCP elicitations on the approval card.t3 upasks how other devices reach the machine; short commands liket3 --versionno longer load the whole server; the Intel Mac installer exits with a clear error; desktop prepares a server bytecode cache in native builds.perfcommits across server reads, git status, ACP terminal output, and web re-render isolation.fixcommits across web (192), server (152), mobile (65), desktop (33), and the rest.Conflict resolutions
27 files conflicted. The main decisions:
allow_this_time,allow_once_only) in both the Claude and Codex matchers, and its rule that elicitation cards have no feedback field.PULL_REQUEST_LINKING_INSTRUCTIONSand the newpullRequestLinkingflag. OpenCode now keeps the watch flag on the turn, since upstream builds the system prompt per prompt.threadDetailReffix replaces the fork's equivalent fix (fix(web): new worktree conversations no longer stay empty #47) and its test.gitPaths. Took upstream's closest-remote merge base and itsnode --testsplit.Waiting · <label>format, with the fork's always-waiting state.ManagedPolicy,BootService.stop,repointLauncher({ baseDir })) and refreshedagentContextBudget.json. The fork's tool-list overhead is the same +121 bytes as before.Verification
After merging, run the workflow-disable step from
docs/operations/fork-server.md. This update adds no new upstream workflows, but it's a quick check.🤖 Generated with Claude Code