Fix Claude Code Review CI quota exhaustion#1353
Merged
sbryngelson merged 5 commits intoMFlowCode:masterfrom Apr 4, 2026
Merged
Fix Claude Code Review CI quota exhaustion#1353sbryngelson merged 5 commits intoMFlowCode:masterfrom
sbryngelson merged 5 commits intoMFlowCode:masterfrom
Conversation
Remove the synchronize event trigger that was causing the review to run on every push to a PR branch, exhausting the API quota. Also remove all incremental review mode logic (now dead code) and filter out tests/ directory from review inputs since golden files don't need code review.
|
Claude Code Review Head SHA: 8cf4f11 Files changed:
Findings:
|
Contributor
Review Summary by QodoFix Claude Code Review CI quota exhaustion and streamline workflow
WalkthroughsDescription• Remove synchronize event trigger causing quota exhaustion • Eliminate incremental review mode logic (now dead code) • Filter tests/ directory from review inputs • Prioritize src/ files for context and increase file size cap • Allow Claude review to soft-fail without blocking CI Diagramflowchart LR
A["Remove synchronize trigger"] --> B["Eliminate incremental mode"]
B --> C["Filter tests/ directory"]
C --> D["Prioritize src/ context"]
D --> E["Soft-fail on error"]
E --> F["Simplified review prompt"]
File Changes1. .github/workflows/claude-code-review.yml
|
Contributor
Code Review by Qodo
1. Rename diff wrongly dropped
|
Contributor
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughThis pull request modifies the GitHub Actions workflow for code review by simplifying its trigger logic and review mechanism. The workflow's 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. Comment |
![qodo-code-review[bot]](https://avatars.githubusercontent.com/in/484649?s=60&v=4){kind=small}
Comment on lines
+116
to
+120
| awk ' | ||
| /^diff --git a\/tests\// { skip=1; next } | ||
| /^diff --git / { skip=0 } | ||
| !skip { print } | ||
| ' .claude-review/review.diff.raw > .claude-review/review.diff |
Contributor
There was a problem hiding this comment.
1. Rename diff wrongly dropped 🐞 Bug ≡ Correctness
The AWK filter strips any diff whose *old* path starts with tests/ (diff --git a/tests/...), which will incorrectly drop diffs for renames/moves from tests/ to non-tests/ paths (e.g., tests/foo -> src/foo). This can cause the review to miss real (non-test) changes while changed_files.txt still includes the new non-test path.
Agent Prompt
### Issue description
The AWK rule that strips `tests/` diffs matches `diff --git a/tests/...`, which incorrectly removes diffs for files renamed/moved from `tests/` to non-`tests/` paths.
### Issue Context
`changed_files.txt` is filtered based on the post-change path list (`.files[].path`), so the diff filtering should key off the *new* path (`b/...`) to stay consistent.
### Fix Focus Areas
- .github/workflows/claude-code-review.yml[113-120]
### Suggested change
Update the AWK logic to set `skip` when the `diff --git` header indicates the **new** path is under `tests/` (e.g., match `b/tests/`). For example:
- On `^diff --git ` lines, compute `skip = ($0 ~ /^diff --git a\/[^ ]+ b\/tests\//)`
- Print lines only when `skip==0`
This keeps renames like `tests/foo -> src/foo` in the diff while excluding files whose new location is `tests/`.
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
Comment on lines
175
to
177
| - name: Run Claude Code Review | ||
| continue-on-error: true | ||
| uses: anthropics/claude-code-action@v1 |
Contributor
There was a problem hiding this comment.
2. Hidden claude step failures 🐞 Bug ✧ Quality
continue-on-error: true on the Claude step prevents the job from failing, so the existing `if: failure()` fallback summary will never run and Claude outages/quota failures become silent. This can leave PRs appearing “successfully reviewed” while no review is published.
Agent Prompt
### Issue description
`continue-on-error: true` makes Claude failures non-fatal, but then `if: failure()` never triggers, so failures are not surfaced anywhere.
### Issue Context
You likely want to avoid blocking CI, but still need visibility when the review step fails (quota/auth/outage).
### Fix Focus Areas
- .github/workflows/claude-code-review.yml[175-177]
- .github/workflows/claude-code-review.yml[266-315]
### Suggested change
1) Give the Claude step an `id` (e.g., `id: claude`).
2) Add a follow-up step that runs **always** (`if: always()`) and checks `steps.claude.outcome` / `steps.claude.conclusion`.
3) If it failed, write a clear message to `$GITHUB_STEP_SUMMARY` (and/or post/update the primary comment with an error marker) so failures are visible even though the job stays green.
ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
synchronizeevent trigger that caused the Claude review to run on every push to a PR branch, exhausting the API quota (You've hit your limiterrors)synchronizeremoval)tests/directory from review inputs — golden files don't need code reviewTest plan
tests/files produce an empty review