
pr-inline-review
by stacklok
ToolHive makes deploying MCP servers easy, secure and fun
Use Cases
Kubernetes Operations
Automate Kubernetes cluster management and deployment. Using pr-inline-review.
MCP Server Integration
AI tool integration using Model Context Protocol. Using pr-inline-review.
API Integration
Easily build API integrations with external services.
SKILL.md
name: pr-inline-review description: Submit inline review comments to GitHub Pull Requests using the GitHub CLI, with support for inline code suggestions.
PR Inline Review
Submit inline review comments to GitHub Pull Requests using the GitHub CLI, with support for inline code suggestions.
Instructions
You are helping submit inline review comments to a GitHub Pull Request with optional inline code suggestions.
Prerequisites
- GitHub CLI (
gh) must be installed and authenticated - User must have write access to the repository
- PR must exist and be open
Workflow
-
Collect findings: The user will provide you with:
- Repository owner and name (or let you detect from current directory)
- PR number
- A list of findings, each containing:
- File path (relative to repo root)
- Line number
- Comment body/description
- (Optional) Suggested fix if it's a simple change
-
Read current content: If providing suggestions, use the Read tool to see the exact current content
-
Create review JSON: Build a JSON structure at
/tmp/pr-review-comments.json:{ "body": "Overall review summary", "event": "COMMENT", "comments": [ { "path": "path/to/file.ext", "line": 123, "body": "Comment text with optional suggestion" } ] } -
Submit review: Use GitHub CLI:
gh api -X POST repos/{owner}/{repo}/pulls/{pr_number}/reviews --input /tmp/pr-review-comments.json -
Return URL: Extract and return the review URL from the response
JSON Structure
Top-level fields
body(required): Overall review summaryevent(required):"COMMENT","APPROVE", or"REQUEST_CHANGES"comments(required): Array of comment objects
Comment object fields
path(required): File path relative to repository rootline(required): Line number (positive integer)body(required): Comment text (supports markdown)
Inline Code Suggestions
GitHub supports inline code suggestions that users can commit directly from the PR UI.
When to Use Suggestions
✅ Good candidates:
- Fixing typos or incorrect file paths
- Correcting simple syntax errors
- Updating version numbers or constants
- Renaming variables or functions
- Fixing formatting or indentation
- Adding missing content
❌ Not suitable:
- Complex logic changes requiring multiple files
- Changes that need testing or validation
- Architectural changes requiring discussion
- Changes requiring user decision/context
Suggestion Syntax
Include a suggestion block in the comment body:
Single-line:
Description of the issue.
```suggestion
corrected line of code
```
Evidence: reference
Multi-line:
Description of the issue.
```suggestion
first corrected line
second corrected line
third corrected line
```
Evidence: reference
Best Practices
Comment Writing
- Be specific with line numbers and file paths
- Provide evidence (link to code/documentation)
- Be constructive - suggest fixes, not just problems
- Use markdown formatting for clarity
- Include context explaining why it's an issue
When Including Suggestions
- Read the current line(s) using Read tool first
- Provide exact replacement text
- Match existing formatting and style
- Verify syntax is correct
- One suggestion block per comment
Review Strategy
- Group related findings into a single review
- Put simple fixes with suggestions first
- Use appropriate event type
- Write clear summary in
body - Use TodoWrite to track progress
Error Handling
- 401 Unauthorized: Run
gh auth login - 404 Not Found: Verify PR number and repo access
- 422 Unprocessable Entity: Check JSON format
- Invalid line number: Ensure line exists at PR's commit
Output Format
Report after submission:
- Review ID and URL
- Number of comments submitted
- Number with suggestions
- PR title and number
See Also
- EXAMPLES.md - Common use cases and examples
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 1000以上
1ヶ月以内に更新
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon

