Skip to content

fix(fork): backport file safety and Codex reliability updates - #21

Merged
unlockdigital merged 6 commits into
unlock-digital/mainfrom
fix/upstream-maintenance-20261114
Nov 14, 2026
Merged

unlockdigital merged 6 commits into
unlock-digital/mainfrom
fix/upstream-maintenance-20261114

Conversation

@unlockdigital

@unlockdigital unlockdigital commented Nov 14, 2026 •

Copy link
Copy Markdown
Owner

Legacy-encoded workspace files could be corrupted on autosave, project deletion could discard worktree cleanup policy, a hung Codex initialization could block later runs, slow valid usage replies were discarded, and pooled quota summaries counted supported subscription tiers equally.

Backport upstream pingdotgg#36545, pingdotgg#36649, pingdotgg#26555, pingdotgg#36189, and pingdotgg#35427 in five focused commits with upstream authorship and commit references. Preserve the fork's account discovery, optional native usage, and weekly quota suppression. Use Map lookups for plan capacities so unknown plan labels consistently fall back to equal weights, and explain the weighting in the usage guide.

Release smoke retains the reviewed lockfile while refreshing package versions, matching the release workflow and avoiding incompatible transitive peers selected by a fresh resolution.

Validation:

  • Reproduced all five original bugs on the published fork: seven failing regression cases, then all nine selected cases passing after the backports.

  • Focused service, storage cleanup, Codex adapter/account, and shared quota suites: 530 tests passed.

  • Changed-file checks: format, lint, both knip checks, GraphQL generation, 35 affected package typechecks, and 364 tests passed.

  • Isolated source server smoke: startup with disabled providers and temporary state, environment discovery HTTP 200, and shutdown through the captured process handle.

  • Release smoke: reproduced the incompatible unenv peer failure, then passed with CI settings after correcting lockfile handling.

Implemented by GPT-6.1 Sol with max reasoning through the Codex harness in T3 Code.

StiensWout and others added 5 commits November 14, 2026 13:08
…tgg#36545)

Co-authored-by: Julius Marminge <julius0216@outlook.com>

(cherry picked from commit 5abf383)
…gdotgg#36649)

Co-authored-by: Julius Marminge <julius0216@outlook.com>

(cherry picked from commit 071ae4e)

Adapted the import context to the fork while retaining its project brief and deletion behavior.
…ex run (pingdotgg#26555)

Co-authored-by: Julius Marminge <julius0216@outlook.com>

(cherry picked from commit f3b43ef)

Assert completion at the virtual deadline so regressions fail without waiting for a wall-clock timeout.
…ropped (pingdotgg#36189)

(cherry picked from commit 8610b31)

Retained the fork's optional native usage and account discovery paths while removing the redundant usage timeout.
…dotgg#35427)

(cherry picked from commit 0b21a82)

Use Map lookups so every unknown plan label falls back to equal weights. Cover inherited object property names and preserve weekly exhaustion/refill handling.
@github-actions github-actions Bot added vouch:trusted PR author is trusted by repo permissions or the VOUCHED list. size:L labels Nov 14, 2026
@github-actions

github-actions Bot commented Nov 14, 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 21.5 KiB 21.5 KiB 0 B (0.0%) 24.4 KiB ✅
Codex Live turn messages 1 1 0 (0.0%) 8 ✅
Claude Total thread wire 5.0 KiB 5.0 KiB 0 B (0.0%) 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 1.2 KiB 1.2 KiB 0 B (0.0%) 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 1 1 0 (0.0%) 8 ✅

Baseline: a0da115 · PR result: 4cc57d1 · 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.

Match the release workflow when refreshing package versions so fresh transitive dependency resolution cannot replace the compatible unenv peer set. Reproduced the failure before this change and passed release smoke with CI settings after it.
@unlockdigital
unlockdigital merged commit 059d8e0 into unlock-digital/main Nov 14, 2026
29 checks passed
@unlockdigital
unlockdigital deleted the fix/upstream-maintenance-20261114 branch November 14, 2026 22:22
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

size:L 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.

4 participants