anthropics / anthropics/claude-code-action
v1.0.5 runs but can't pick up the PR context nor post the output as a comment on the PR itself.
- Lingua principale
- TypeScript
- Stelle
- 8.9k
- Fork
- 2.1k
- Merge medio
- 3g 9h
- PR unite (30g)
- 10
Descrizione
**Describe the bug**
GH Claude v1.0.5 actions runs as a result of tagging `@claude` in a comment with an ask to review the PR. However, it the action doesn't seem able to pick up the PR context, and won't post anything on the PR itself as a comment.
What am I missing?
**To Reproduce**
Steps to reproduce the behavior:
1. Go to open PR
2. Tag `@claude` in a comment with an ask to review PR
3. GH Claude action runs
4. No output / comment gets posted on PR
**Expected behavior**
GH Claude action runs successfully, picks up the PR context, posts the output as a comment on the PR itself.
**Screenshots**
Triggering Claude:
Claude action summary:
**Workflow yml file**
```
name: Claude Code Review
on:
issue_comment:
types: [created]
pull_request_review_comment:
types: [created]
pull_request_review:
types: [submitted]
jobs:
claude-review:
if: |
(github.event_name == 'issue_comment' && github.event.issue.pull_request && contains(github.event.comment.body, '@claude')) ||
(github.event_name == 'pull_request_review_comment' && contains(github.event.comment.body, '@claude')) ||
(github.event_name == 'pull_request_review' && contains(github.event.review.body, '@claude'))
runs-on: ubuntu-latest
permissions:
contents: read
pull-requests: write
issues: write
id-token: write
steps:
- name: Checkout repository
uses: actions/checkout@v4
with:
fetch-depth: 1
- name: Run Claude Code Review
uses: anthropics/claude-code-action@v1.0.5
with:
anthropic_api_key: ${{ secrets.ANTHROPIC_API_KEY }}
# Minimal, actionable review prompt - no positive feedback
prompt: |
Review this PR and provide ONLY specific, actionable feedback. Do not provide:
- Positive feedback, praise, or "LGTM" comments
- General suggestions or best practices
- Formatting or style nitpicks
- Explanations of what the code does
ONLY comment if you find:
- Security vulnerabilities with exact fix
- Bugs with specific line numbers and solutions
- Breaking changes that need addressing
- Missing error handling with exact implementation
- Performance issues with specific optimization
- Code quality issues like violations of established patterns, maintainability concerns, and significant readability issues
Format: "[file:line]" for line-specific issues, "[file]" for file-level issues. Always include exact fix.
If nothing actionable found, return "Nothing actionable found 👍"
# Use sticky comments to make Claude reuse the same comment on subsequent pushes to the same PR
use_sticky_comment: true
# Tools for comprehensive PR review
claude_args: |
--append-system-prompt "Please refer to the project's CLAUDE.md file and .claude/agents/ documentation for:
- Architecture patterns and conventions to follow
- Testing patterns and requirements
- Code style guidelines and established practices
- Technology stack and framework usage
Pay special attention to violations of established patterns documented in these files."
--allowedTools "mcp__github_inline_comment__create_inline_comment,Bash(gh pr comment:*),Bash(gh pr diff:*),Bash(gh pr view:*)"
```
**API Provider**
[ x] Anthropic First-Party API (default)
[ ] AWS Bedrock
[ ] GCP Vertex
Guida per i contributori
Apri la guida per i contributori
Valutazione
Questa issue non è ancora stata valutata.