fix: close trace HTTP test response and listener resources

[seonghobae]

Scope and ownership

Stacked on #1140 at 38c0603. This preserves its production response-lifecycle delta and adds only trace HTTP test-resource ownership plus documentation. Production authentication, routing, retries, and response bodies are unchanged.

The test client's urllib HTTPError is separate from the server's serialized authorization failure. Both test request helpers now close that response after decoding while preserving primary decoding errors if cleanup raises OSError. All 23 test listener teardown paths close their sockets after shutdown/join. No warning suppression or global cleanup is introduced.

Evidence

• Parent trace singleton: 1 failed under warnings-as-errors.
• RED checkpoint a9de9de: 8 explicit ownership regressions failed.
• Tested source e885621: trace strict 31 passed; related strict 126 passed; full default 3670 passed, 2 skipped.
• Full strict remains RED: 1173 failed, 2494 passed, 2 skipped, 13 errors. No target-module failure appears in the aggregate failure list; this does not establish the remaining failures' causes or a causal percentage improvement.
• Independent read-only review found no actionable finding and reproduced 31 strict passes. This is not required GitHub approval.
• Final head b9ad721 adds documentation only after tested e885621; tests are not relabeled as executed at this later head.

Visual inspection and limits

At b9ad721, the changed runbook, Gap baseline, AGENTS and CLAUDE sections were rendered in a real browser at 1265 x 712, English, and screenshots directly inspected. New paragraphs and the complete evidence table were readable without overlap or horizontal clipping. This is a bounded documentation inspection, not full product UI, responsive, or multilingual acceptance.

Exact-head commands, logs and limitations are in docs/doctoring/http_test_resource_lifecycle.md. Cleanup injection covers OSError after underlying close, not arbitrary exception classes or a close that cannot complete. No measured customer accuracy or decision-latency improvement is claimed. Remains Draft pending full strict remediation, required hosted checks and independent review; no merge or deployment is requested.

[coderabbitai]

Warning

Review limit reached

Next included review available in 16 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: c7671162-f9e2-46cf-9e90-186c14e64def

📥 Commits

Reviewing files that changed from the base of the PR and between 957b174 and 3a821bf.

📒 Files selected for processing (5)

• AGENTS.md
• CLAUDE.md
• docs/doctoring/http_test_resource_lifecycle.md
• docs/product-technical-gap-baseline.md
• tests/test_chat_include_orchestration_trace_http_honesty.py

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[seonghobae]

Restacked on the updated base branch (a675e7a9; merged codex/stream-error-resource-lifecycle-20260912 after #1140 took current main — clean, no conflicts). Local: tests/test_chat_include_orchestration_trace_http_honesty.py 31 passed, api-contract/self-check 9 passed, interrogate 100%. Still draft; carried by this session since the Codex writer went inactive 2026-09-13.

🤖 Addressed by Claude Code