merge: first upstream batch for completion state and background-thread shortcuts

[Hylouis233]

Merged: first upstream integration batch

Merged into the owner's main on 2026-10-15 with a history-preserving merge commit:

• Merge commit: 5cd52d7c79d323c6b8fbd79c0f3f3ee041a2220a.
• Verified candidate head: 188885079eab8b7ef72bfda1f78fc161a1753beb.
• Candidate tree: 41c9e1868e1314b7d31f30bf7a621c9d03d56f50.
• Previous fork baseline: a02ffa38e8401275da3acf76436d8dc99367fceb.
• Actual common ancestor: a5b34b25378fbbf90ce36ee101a11528ccf11c7e.

This supersedes the earlier draft/preparation status; the merge has now actually occurred.

Included upstream changes

The batch integrates two upstream commits while preserving the fork tree outside the 14 changed files:

• f68e24fb419c985269cd0acb841e203af1419c43: completion visibility when a thread leaves a shell command or development server running (upstream fix(web): a thread that left a shell running shows its unseen completion pingdotgg/t3code#14910).
• 263097a8da038a143f2b7b7817250f0ad390ea32: restore mod+Enter background creation for new threads, with one-time default-keybinding migration retaining custom bindings and later removals (upstream fix(web): mod+enter starts a new thread in the background again pingdotgg/t3code#15060).

The candidate's reconstructed tree was required to match the exact three-way Git merge tree. Changed-file manifests and SHA-256/Git blob identities were checked. The temporary preparation workflow is absent from the final tree; history was not force-rewritten.

Verified before merge

• Immutable candidate preparation: successful, including the exact 14-file allowlist and tree equality.
• Application CI: successful on the candidate head.
• Fork maintenance checks: successful; job inspection confirmed executed server tests, type checking, desktop/web builds and maintenance checks rather than an empty run.
• Upstream sync invariant: successful.
• No unresolved review threads were returned immediately before merge.

Git mergeability alone was not treated as an application-test pass. Live desktop/production acceptance is not claimed.

Preserved boundaries and remaining synchronization

MCode/DSH adapters, credentials, permission policy, runtime-ownership fixes, dependencies/lockfile, CI/publishing restrictions, repository settings and .repos are unchanged relative to the fork baseline. No production deployment, package release, paid provider request or local user-data modification was performed.

#11 remains open. The full pinned upstream target e6bbfdb9ffe8334bed1c9b992f828f04d30db7ea previously produced 21 conflicted files and a larger provider/session migration. This first merged batch does not complete that full synchronization. Remaining candidates must be recomputed against the new main and verified without dropping fork-specific capabilities.

Authored/integrated with GPT-6 Astra Pro in ChatGPT, with preparation and validation evidence from this fork's GitHub Actions.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	4.9 KiB	4.9 KiB	0 B (0.0%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.7 KiB	3.7 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	20.4 KiB	20.4 KiB	0 B (0.0%)	29.3 KiB	✅
Codex	Live turn messages	2	2	0 (0.0%)	8	✅
Claude	Total thread wire	4.9 KiB	4.9 KiB	−41 B (−0.8%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.7 KiB	3.7 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	−41 B (−3.4%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	20.8 KiB	20.7 KiB	−41 B (−0.2%)	29.3 KiB	✅
Claude	Live turn messages	2	1	−1 (−50.0%)	8	✅

Baseline: a02ffa3 · PR result: 1888850 · 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: 106.1 KiB
• Claude decoded thread snapshot: 106.4 KiB

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

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.