fix: prevent duplicate PR review comments by tracking existing comments

This commit is contained in:
dayuan.jiang
2025-11-29 12:51:34 +09:00
parent 28258d19ca
commit 2f01f016d9

View File

@@ -77,16 +77,17 @@ jobs:
- Vercel AI SDK (streamText, useChat, tool calling)
- Multiple AI providers: Bedrock, Anthropic, OpenAI, Google, Azure, OpenRouter, Ollama
First, check previous review comments from github-actions bot using `gh api repos/${{ github.repository }}/pulls/${{ github.event.pull_request.number }}/comments`.
For each previous comment:
- If the issue is fixed in the current code, resolve the comment thread using:
STEP 1: Check existing comments to avoid duplicates.
Run: `gh api repos/${{ github.repository }}/pulls/${{ github.event.pull_request.number }}/comments`
Build a list of files and line numbers that already have comments. For each existing comment:
- If the issue is FIXED in current code, resolve the thread using:
`gh api graphql -f query='mutation { resolveReviewThread(input: {threadId: "THREAD_ID"}) { thread { isResolved } } }'`
Get the thread ID from the comment's node_id field.
- If the issue still exists, leave it alone
- If the issue still exists, remember this file:line - DO NOT create a new comment for it
Then review the current diff for NEW issues only:
STEP 2: Review the diff for issues, but SKIP any file:line that already has a comment.
Review this PR for ONLY these issues:
Review this PR for these issues (report ALL that apply):
1. Bugs that would cause runtime errors or broken functionality
2. Security issues (exposed secrets, API key leaks)
3. AI SDK misuse - specifically check for:
@@ -96,18 +97,22 @@ jobs:
- Tool definitions: Must use Zod schemas for inputSchema
- Status handling: Check status (submitted/streaming/ready/error) before actions
- Stream cleanup: Call stop() when aborting streams
4. Unrelated changes that should be in separate PRs (scope creep)
5. Suspicious .gitignore additions or accidentally committed files
6. UI/UX inconsistencies (e.g., alignment issues)
When reviewing AI SDK usage, fetch https://ai-sdk.dev/docs/ to verify correct patterns.
Key doc pages: /docs/ai-sdk-ui/chatbot, /docs/ai-sdk-core/generating-text, /docs/ai-sdk-core/tools-and-tool-calling
DO NOT comment on:
- Performance optimizations
- Code style or formatting
- "Best practices" that don't affect functionality
- Type safety improvements
- Error handling additions
- Minor performance optimizations
- Code style preferences (unless clearly wrong)
- Type annotations that don't affect functionality
Use `mcp__github_inline_comment__create_inline_comment` for inline comments.
Be very selective - if there are no real bugs, just say "LGTM" in a PR comment.
IMPORTANT:
- NEVER create a comment on a file:line that already has a comment - this causes duplicates
- For each NEW issue, use `mcp__github_inline_comment__create_inline_comment` to comment on the specific line
- ALWAYS include a suggested fix using GitHub's suggestion syntax: ```suggestion\n<fixed code>\n```
- Only say "LGTM" if there are truly ZERO new issues to report
claude_args: |
--allowedTools "mcp__github_inline_comment__create_inline_comment,Bash(gh pr comment:*),Bash(gh pr diff:*),Bash(gh pr view:*),Bash(gh api:*),WebFetch(domain:ai-sdk.dev)"
--allowedTools "mcp__github_inline_comment__create_inline_comment,Bash(gh pr diff:*),Bash(gh pr view:*),Bash(gh api:*),WebFetch(domain:ai-sdk.dev)"