docs: trace conda_avail re-wiring insertion point; decision to defer implementation - #393
Conversation
Completes the scoping this repo's own doc explicitly asked for before any further conda_avail re-wiring: confirmed selfapps_envsmoke.ps1 is the genuine first real-install candidate under HP_FORCE_CONDA_ONLY=1 (single/entry/isolation/envname all set HP_CI_SKIP_ENV; selftests.ps1 replays a log; selfapps_size.ps1 never bootstraps at all). Deliberately not implementing the re-wiring itself in this pass -- see the new paragraph for why a third attempt at this exact mechanism (already the source of two real, subtle bugs in quick succession) needs explicit owner sign-off rather than autonomous action. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_015xbWLPbiaKVsobB9FZy8kS
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
📜 Recent review details⏰ Context from checks skipped due to timeout. (9)
🧰 Additional context used📓 Path-based instructions (2)**/*📄 CodeRabbit inference engine (AGENTS.md)
Files:
**/*.{md,bat,cmd}📄 CodeRabbit inference engine (AGENTS.md)
Files:
🔇 Additional comments (1)
📝 WalkthroughSummary by CodeRabbit
WalkthroughCLAUDE.md item 7 now documents the CI step-order trace, identifies ChangesCI backlog clarification
Estimated code review effort: 1 (Trivial) | ~2 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Summary
Completes the scoping pass CLAUDE.md's own Active Backlog explicitly asked for before any further
conda_availwork: "this needs a proper trace, not a guess, before being reintroduced."batch-check.yml's fullselftestjob step order from the top, checking each step for anHP_CI_SKIP_ENVoverride or a static-only check that never invokesrun_setup.bat.selfapps_single.ps1/selfapps_entry.ps1/selfapps_isolation.ps1all setHP_CI_SKIP_ENV: '1'at the step level (previously only cited, now directly confirmed); found a fourth candidate not previously named (selfapps_envname.ps1, sets it internally); confirmedselftests.ps1replays a captured log rather than re-bootstrapping, andselfapps_size.ps1is a static byte-size check.selfapps_envsmoke.ps1("Self-test: real env smoke (CI-only)") -- its own script comment says "FULL bootstrap here: do NOT set HP_CI_SKIP_ENV," it runs unrestricted by lane, and it sits immediately before the firstreal/conda-full onlyself-test block.conda_avail, gating the two must-not-regress lanes) has already produced two real, subtle bugs in quick succession -- the original premature-gating bug (ci: fix conda_avail circular gate silently skipping conda-full self-tests #391) and a distinct wording slip CodeRabbit caught on the PR that reverted it (docs/tools: address PR #391 CodeRabbit findings; add advisory MD029 markdownlint check #392). A third attempt at the same fragile spot, even with the trace now in hand, is treated as architecturally significant enough to need explicit owner sign-off rather than autonomous action, per this repo's own standing escalation policy. The full implementation shape (move the existing unusedconda_availstep to right afterselfapps_envsmoke.ps1, re-point the same 36if:conditions) is documented and ready if/when approved.Doc-only change, zero
run_setup.bat/workflow-YAML risk.Test plan
python -m compileall -q .python tools/check_delimiters.py run_setup.batmarkdownlint-cli2 CLAUDE.md(clean, only the expected permanent baseline finding)python -m yamllint .github/workflows/actionlint -oneline .github/workflows/*.ymlpython -m pytest tests/test_*.py -q(437 passed, 2 skipped)tools/run_sanity_sweep.shrun, all checks OKClaude-Session: https://claude.ai/code/session_015xbWLPbiaKVsobB9FZy8kS
Generated by Claude Code