Repository navigation
chore: integrate T3 Code v0.1.2 and retain upstream test pruning - #443
Merged
Merged
Conversation
…e (#22570) Co-authored-by: Claude Opus 5.5 <[email protected]>
…#22557) Co-authored-by: Claude Opus 5.5 <[email protected]>
…#22575) Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
…event-driven (#22565) Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
…ing parent (#22581) Co-authored-by: syberia <[email protected]> Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
…592) Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
… (#22599) Co-authored-by: Claude Opus 5.5 <[email protected]>
…556) Co-authored-by: Claude Opus 5.5 <[email protected]>
…rts (#22602) Co-authored-by: Claude Fable 5.1 <[email protected]>
…559) Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
…23124) Co-authored-by: Claude Opus 5.5 <[email protected]>
…s (#23389) Co-authored-by: PR Batch Tester <[email protected]> Co-authored-by: Claude Opus 5.5 <[email protected]>
…desktop updates (#23113) Co-authored-by: PR Batch Tester <[email protected]> Co-authored-by: Claude Opus 5.5 <[email protected]>
…s (#23454) Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: Claude Opus 5.5 <[email protected]>
…(#23135) Co-authored-by: PR Batch Tester <[email protected]> Co-authored-by: Claude Opus 5.5 <[email protected]>
… columns (#23502) Co-authored-by: PR Batch Tester <[email protected]> Co-authored-by: Claude Opus 5.5 <[email protected]>
Co-authored-by: GPT-6.1 Sol <[email protected]>
Co-authored-by: codex <[email protected]>
…(#23519) Co-authored-by: Claude Opus 5.5 <[email protected]>
Contributor
WalkthroughWarning Review details and warnings were omitted to fit the comment limit. |
github-actions
Bot
added
size:XXL
Contributor
Thread transfer impact✅ Thread transfer remains within every enforced ceiling.
Baseline: Scenario and decoded snapshot size10 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.
Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed. |
dbalders
marked this pull request as ready for review
October 25, 2026 23:18
dbalders
commented
Oct 25, 2026
dbalders
left a comment
dbalders
left a comment
Owner
Author
There was a problem hiding this comment.
TritonAI Review unavailable
The review could not finish. No code verdict was produced. Reason: PermanentReviewError: PR evidence exceeds one focused review context; no review verdict was produced. Split the PR or use a reviewer with a larger context. Automatic chunk reviews are disabled.
Completed review passes remain cached. A maintainer can comment /triton-review to retry; service health and logs contain the failure state.
Owner
Author
Cross-model review unavailableImplementation (PR footer): GPT-6.1 Sol via Codex in TritonAI Harness The review did not finish, so no verdict was produced. Reason: RuntimeError: autoreview published no status (exit 1, reason None) A maintainer can comment Automated cross-model review on ChelsClaw, updated 2026-10-25T23:29:49+00:00. |
dbalders
added a commit
that referenced
this pull request
Oct 26, 2026
) (#444) * test(server): prune low-value fork tests (first pass) Applies the upstream second-pass standard (pingdotgg/t3code#22592) to tests the fork added in apps/server, and drops cases upstream deleted in v0.1.2 that the sync merge kept. Fix regressions are annotated with // regression: comments. Inventories and per-area reports live outside the repo in the task evidence directory. Non-test changes are export-keyword removals for helpers only deleted tests imported, plus removal of the unused CodexPreviewDynamicTools failure-result helper. * test(server): prune low-value fork tests (second pass, two-voter) Runs the upstream first-pass rubric (pingdotgg/t3code#22557) over every test case the fork added in apps/server. Claude and GPT-6.1 Sol voted independently on each of 880 cases; a case is deleted when both agree, a third vote settled 133 disagreements, and cases upstream deleted in v0.1.2 that the sync merge kept are dropped unless both voters kept them. 196 cases and 5 files removed. Twelve helpers only deleted tests used are un-exported; two that nothing calls any more are deleted. * test(server): drop migration upgrade groups emptied by the prune * test(clients): prune low-value fork tests and mark retained regressions * test(clients): drop remaining forwarding and schema literal checks * test(clients): prune low-value fork tests (second pass, two-voter) Runs the upstream first-pass rubric (pingdotgg/t3code#22557) over every test case the fork added outside apps/server. Claude and GPT-6.1 Sol voted independently on each case; a case is deleted when both agree, a third vote settled 233 disagreements, and cases upstream deleted in v0.1.2 that the sync merge kept are dropped unless both voters kept them. 563 cases and 25 files removed. Helpers only deleted tests used are un-exported (30) and three that nothing calls are deleted; the hand-run analytics report script becomes a knip entry. * test: keep three security and managed-policy guards flagged in review Restores the filesystem-permission denials for Project Brief, memory and archive reads, the synchronous-parser timeout kill for untrusted documents, and the managed-skill ownership cases in providerSkillGrouping.
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.
Summary
Integrate T3 Code v0.1.2 (
044f07257c9684ae4543bd1e1ebf9c394e9e3830) while preserving intentional TritonAI behavior and upstream merge ancestry. Retain upstream's test pruning: 465 existing test files removed, 29 added, 436 fewer files overall, and 254,822 net test lines removed relative to the original downstream base.Why
Bring Harness onto the latest stable upstream release and remove the low-value tests upstream retired. Keep meaningful downstream security, migration, managed-policy and behavior coverage through the extracted web/mobile architecture.
Validation
2ea37f5f851c209dafa4973bbadef0a16dfcccfd, including aggregate Check, web/nonserver Test, all five server shards, Windows regression, native static/fingerprint and release-smoke checks. Optional draft preview/review jobs remain skipped. The thread-transfer report matches this head and exact main baseline5bd4bdc; every enforced ceiling passes.preflight:checkspassed, including lint/format, release/process guards, knip and 38/38 typecheck tasks. Desktop build/preload/main-import checks passed 9/9; native iOS build and mobile static checks passed..tgzarchives; the incoming archive and patch Git blobs match the exact upstream release. That provenance check is not a complete binary audit.Verification incomplete: exact-base/candidate visual comparison, live packaged managed-plugin operation, and actual packaged Windows/macOS install/upgrade behavior remain unverified. Android SDK is unavailable on the verification host. Current-head CI passed; the acceptance limits above and incomplete hosted reviews remain explicit.
Risk and Recovery
This is a broad upstream update across server, web, desktop, mobile and the toolchain. Preserve encrypted storage and ownership checks, managed provider policy, downstream attachments/dictation/Briefs/Project Home, and installer-owned updates. Mobile outdated-server messaging remains informational rather than running an upstream remote installer command. Use a deliberate revert plan if necessary; keep the upstream merge ancestry rather than squash-merging this sync.
UI Changes
Candidate desktop after restart and a fresh iOS request:
Release and Deployment Impact
Imports upstream dependencies/native patches, Electron 44.9.3 and Node 26.12.1 / pnpm 12.14.0 / Vite+ 1.3 toolchain changes. Keeps TritonAI release/update ownership and managed composition controls. No release, deployment or App Store review action is part of this PR.
Upstream Alignment
Upstream v0.1.2 remains an ancestor through a merge commit. Start from upstream's pruned test structure and extracted components, then retain meaningful complete downstream behavior cases. Intentional divergence remains explicit for TritonAI branding, managed models, UCSD integrations, security and release control.
Checklist
Implementation: GPT-6.1 Sol via Codex in TritonAI Harness
Note
Integrate T3 Code v0.1.2 upstream changes and retain pruned upstream tests
tsc-rsto the standard TypeScript compiler across packages, and updates workspace dependency versions in pnpm-workspace.yamlMacroscope summarized 2ea37f5.