fix(prime): retain safe session closure diagnostics

[rynfar]

Unexpected Prime session closures were reduced to a generic “stopped before sending a final response” error, leaving no safe way to distinguish provider closure from internal proof, capacity, snapshot, or MCP recovery failures.

Record a bounded internal closure classification and generation metadata in provider event logs. Ignore classifications supplied by raw provider events, retain private-error redaction, and preserve existing terminal behavior and recovery fences. This is diagnostic instrumentation: the historical incident could not be reproduced or assigned a proven root cause from the available logs.

Validation: 574 tests across the four focused Prime daemon runtime, adapter, and event suites; scoped lint; server typecheck (existing Effect suggestions only); diff whitespace check. Includes regression coverage for private-error suppression, untrusted diagnostic fields, and internally synthesized recovery failures. No client or wire changes.

Implemented through Pylon delegation by Antigravity (gemini-3.8-flash-high); reviewed, corrected, and verified by GPT-6 in the Codex harness.

---

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

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
pylon-marketing	Ready	Preview	Sep 17, 2026 7:15pm UTC

[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	13.9 KiB	14.0 KiB	+30 B (+0.2%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.2 KiB	7.2 KiB	+9 B (+0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.7 KiB	6.8 KiB	+21 B (+0.3%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	58.0 KiB	58.0 KiB	+44 B (+0.1%)	66.4 KiB	✅
Codex	Live turn messages	9	10	+1 (+11.1%)	21	✅
Claude	Total thread wire	14.0 KiB	14.0 KiB	+12 B (+0.1%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.2 KiB	7.2 KiB	−7 B (−0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.8 KiB	6.8 KiB	+19 B (+0.3%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	58.9 KiB	58.9 KiB	+44 B (+0.1%)	66.4 KiB	✅
Claude	Live turn messages	9	10	+1 (+11.1%)	21	✅

Baseline: ad33250 · PR result: 3a92c40 · 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: 115.6 KiB
• Claude decoded thread snapshot: 116.3 KiB

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