[yordis]

• Upstream deleted roughly 11k low-signal tests and added lint rules against per-test timeouts, wall-clock reads, real-time waits, and polling loops. Every fork test file that conflicted was reduced to the cases that cover a ledger divergence, so the fork does not carry back the test volume upstream removed.
• The release workflow keeps the fork's GitHub-hosted runners and SHA-pinned actions because the org allowlist rejects unpinned and unlisted actions. The Windows x64 job takes upstream's build-on-Linux and verify-on-Windows shape with a GitHub-hosted verify runner instead of Blacksmith.
• The model picker keeps the remembered opening tab (ledger 0032) on top of upstream's new component structure, and takes upstream's de-export of the adjacent-provider helper.
• The RPC instrumentation test was rewritten to a single case proving server spans join the client's trace, since upstream replaced the surrounding fixtures the previous tests depended on.
• Tests that upstream now covers elsewhere were taken from upstream rather than restored: GitManager hook output is covered by the VCS driver core test, the Mermaid warmup in the timeline test is no longer needed, and the OTLP endpoint permission case was superseded by the new secret-source permission test.
• The generated route tree was accepted from the merge instead of restored from the fork, per the sync skill's generated-files rule, so the merge base stops carrying the stale header.
• Full typecheck, lint, and test runs pass locally except two upstream desktop icon tests that write two file names differing only by case to a real directory, which collide on a case-insensitive macOS volume; those tests are unchanged from upstream and run on Linux in CI.

---

^{Need help on this PR? Tag @codesmith-bot with what you need.}

[cursor]

This PR is too large for Bugbot to review. It changes 71,931 lines and 3,849,405 characters. Split the change into smaller pull requests to get a review.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ The exact PR base did not have a successful artifact. Baseline uses the latest successful main measurement shown below.

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	20.9 KiB	20.9 KiB	0 B (0.0%)	29.3 KiB	✅
Codex	Live turn messages	2	2	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	21.2 KiB	21.2 KiB	0 B (0.0%)	29.3 KiB	✅
Claude	Live turn messages	2	2	0 (0.0%)	8	✅

Baseline: 09f1583 · PR result: bfdc584 · 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.