[unlockdigital]

Integrate official v0.1.9 into the maintained public fork and prepare the next isolated macOS preview. Preserve upstream ancestry and supported public customizations: Home, account usage, privacy routing, project briefs, scoped thread links, and Upstream Updates.

Reconcile Effect 4 and provider SDK changes, retain security floors, and retarget the Claude/Cursor embedded security patches. Restore Upstream Updates under the current Settings layout and keep billing links in account details. Increase the size-label output buffer for this large upstream import. Record pingdotgg#32959 as upstreamed; v0.1.9 predates its merge, so the equivalent carried fix remains.

The preview workflow requires the tagged source on the default branch with a successful Check, launches the packaged Mac app against an older synthetic database, and emits source identity and checksums. Application replacement stays manual.

Validation: check:changed v0.1.9 --concurrency 2 passed formatting, lint, both knip checks, patch registration, generated GraphQL checks, all 41 affected package typechecks, 5,116 tests in 229 files (six existing skips), and 92 React Compiler checks. The workflow-only follow-up passed its scoped checks. Canonical CI passed on the merge commit. Browser smoke covered migrated data, a completed free Replay turn with approval, Home, account sorting, and official release refresh. Release CI passed native Mac desktop launch and CLI archive smoke. Downloaded source identity, CLI version, ZIP integrity, and final asset digests/checksums were verified.

Published v0.1.9 Unlock Digital preview from canonical c13dae4d88c4f935f8f394b88d93fdd23555d2f4. The release includes validation, installation guidance, source manifest, and checksums.

The tool catalog grows by 2,027 bytes while provider instruction text falls by about 6.1 KB per provider. The agent-context-increase label accepts the catalog growth inherited from v0.1.9.

Integration candidate before Settings route reconciliation:

Qualified Settings route after reconciliation:

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

[unlockdigital]

The size-label job failed with spawnSync git ENOBUFS: the upstream integration produces 2,639,042 bytes of numstat output, exceeding Node’s default 1 MiB buffer. Reproducing the same diff with a 16 MiB buffer succeeds and classifies 2,658,366 non-test changed lines as size:XXL; that label is applied.

The candidate raises the buffer for both sizing calls. This pull_request_target workflow runs from the existing base branch, so the current failure reflects the old workflow. The required Check and release qualification gates remain enforced.

[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	5.0 KiB	5.0 KiB	−35 B (−0.7%)	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	1.3 KiB	1.2 KiB	−35 B (−2.7%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	21.6 KiB	21.5 KiB	−41 B (−0.2%)	24.4 KiB	✅
Codex	Live turn messages	2	1	−1 (−50.0%)	8	✅
Claude	Total thread wire	5.1 KiB	5.0 KiB	−35 B (−0.7%)	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	1.3 KiB	1.2 KiB	−35 B (−2.7%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.9 KiB	21.9 KiB	−41 B (−0.2%)	24.4 KiB	✅
Claude	Live turn messages	2	1	−1 (−50.0%)	8	✅

Baseline: 5689be8 · PR result: 9efb22a · 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.