Scale PR review sub-agents by diff size

[strawgate]

Summary

• Pre-computes PR size metrics in pr-context and writes pr-size.txt as {N} files, {M} diff lines.
• PR review prompt now uses diff line count to choose fan-out instead of always spawning 3 sub-agents:
  • Under 200 lines: review directly (no sub-agents)
  • 200-800 lines: 2 sub-agents (A→Z, Z→A)
  • Over 800 lines: 3 sub-agents (A→Z, Z→A, largest first)
• Keeps the fan-out decision logic explicit in the PR review prompt while using precomputed pr-size.txt as input.
• Regenerates affected compiled workflow lock files from these prompt/fragment changes.

Supersedes #510 with a simpler approach.

Test plan

• make compile succeeds
• All 32 tests pass
• Verify on a small PR (<200 diff lines) that the agent reviews directly
• Verify on a large PR (>800 diff lines) that 3 sub-agents are spawned

Generated by Update PR Body for issue #512

[coderabbitai]

Warning

Rate limit exceeded

@strawgate has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 4 minutes and 44 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.

📥 Commits

Reviewing files that changed from the base of the PR and between b3b2795 and a74a668.

📒 Files selected for processing (7)

• .github/workflows/gh-aw-fragments/pr-context.md
• .github/workflows/gh-aw-mention-in-pr-by-id.lock.yml
• .github/workflows/gh-aw-mention-in-pr-no-sandbox.lock.yml
• .github/workflows/gh-aw-mention-in-pr.lock.yml
• .github/workflows/gh-aw-pr-review-addresser.lock.yml
• .github/workflows/gh-aw-pr-review.lock.yml
• .github/workflows/gh-aw-pr-review.md

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch feat/pr-review-fanout

---

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

[github-actions]

[MEDIUM] Diff-size fallback misclassifies PRs when full diff fetch fails

DIFF_LINES is derived only from /tmp/pr-context/pr.diff. In the failure path just above, gh pr diff can fail and the script intentionally creates an empty file (: > /tmp/pr-context/pr.diff), which forces DIFF_LINES=0 and classifies the PR as "small" even when the actual patch set is large. This defeats the fan-out logic introduced in this PR and can under-review large PRs; please fall back to counting lines from files.json patches when pr.diff is empty.