chore(summary): drop --fix-first and --held, which nothing reads

[ZhenghuaBao]

Orca-Code-Review — push 1

Severity	Count
P0	0
P1	0
P2	0
P3	0

✅ no blocking findings

summary-comment.mjs parses two flags that nothing downstream reads.

Both survived the cascade that gave them meaning. --held marked a cheap pass withholding a strong review, and --fix-first let the blocking count follow the fix-first set on such a run — because block-on need not contain those severities, and "❌ 0 findings block merge" beside a held tier line contradicts itself.

There is no withholding and no tier line any more, so the count has followed block-on alone for some time. Meanwhile --held was still parsed into opts.held, and --fix-first was still parsed and validated against SEVERITIES, with neither value read anywhere after that.

fix-first the input is untouched

It stops the exhaustive loop early (action.yml), which is a different question from what this summary counts. The header now says so explicitly, so the shared name cannot argue its way back into the count later.

Safe to ship on its own

The argument loop has no else branch, so an unrecognised flag — and the value after it — are skipped one token at a time without shifting the flags that follow. The surviving block-on test now passes --fix-first P0 in the middle of the argv rather than at the end, which turns that property into a regression guard instead of an assumption.

Docs

skills/orca-review-action/references/inputs.md now scopes itself to this action. The hosted GitHub App runs a different review engine, so a setting of the same name can behave differently there — fix-first and exhaustive in particular — and the file says to answer App questions from the console rather than from these rows.

Verification

npm test: 558 pass / 9 fail before this change, and 558 pass / 9 fail after — byte-identical counts. The 9 are a pre-existing CRLF mismatch in the action.yml input assertions on a Windows checkout, unrelated to these files.

🤖 Generated with Claude Code

[ZhenghuaBao]

@codex review

[ZhenghuaBao]

@orcacode-review

[orcacode-review]

🐳 OrcaCode Review

✅ No findings — nothing to flag in this PR. Great work!

_{OrcaCode Review — Route Smarter. Ship Safer. Spend Less.
Engine-reported: 164 calls · 4M tokens · 92% cached}

❤️ Share · Install OrcaCode Review

Free on GitHub — the review runs on your own OrcaRouter key. If it helped, a shout-out goes a long way.

Share: X · Reddit · LinkedIn
Follow: X · Discord · LinkedIn · OrcaRouter