ci: optimize Claude review workflow — enforce skills, reduce token waste

[lklimek]

Summary

Fixes three issues observed in the CI review run for PR #719 (log analysis):

• Remove MCP probe step — the 6-line get_me probe was never actually called (Claude used ToolSearch instead, giving a false positive). Replaced with a single line: "Use MCP tools, fall back to gh CLI on failure."
• Remove unconditional dash-platform skill load — was injecting ~1,500 tokens/turn of Platform SDK docs regardless of PR content. UI/wallet PRs don't need it.
• Enforce Skill tool invocation for check-pr-comments and grumpy-review — Claude was ignoring the skill instructions and doing freestyle reviews inline. Now uses explicit Skill(skill="...") syntax with "do NOT skip" enforcement.

Context

Analysis of PR #719 review run showed:

• 24 turns, $1.30 cost, 3m 37s wall clock
• ~13s wasted on false-positive MCP probe → denial → gh CLI fallback
• ~36K tokens wasted on irrelevant Platform SDK context
• Neither check-pr-comments nor grumpy-review skills were invoked despite prompt instructions

Test plan

• Trigger a review on a simple PR — verify check-pr-comments and grumpy-review skills are actually invoked (visible in log output as Skill tool calls)
• Verify MCP tools work without the probe step (should succeed directly with mcp__plugin_claudius_github in allowedTools from build: fix github mcp access denied #722)
• Verify dash-platform skill is NOT loaded for non-Platform PRs

_{🤖 Co-authored by Claudius the Magnificent AI Agent}

[coderabbitai]

Warning

Rate limit exceeded

@lklimek has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 11 minutes and 52 seconds before requesting another review.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 4502b40c-dcbe-43a7-9ef9-987ce737f973

📥 Commits

Reviewing files that changed from the base of the PR and between caff13f and 699944c.

📒 Files selected for processing (1)

• .github/workflows/claude-code-review.yml

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch fix/ci-review-allowedtools

---

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 and usage tips.}