بنقرة واحدة
review-respond
// Respond to all pending review comments on the current PR — fetch comments, apply fixes, verify accuracy, test, commit, and reply. Use when addressing Copilot reviews, GitHub PR reviews, or any batch of review feedback.
// Respond to all pending review comments on the current PR — fetch comments, apply fixes, verify accuracy, test, commit, and reply. Use when addressing Copilot reviews, GitHub PR reviews, or any batch of review feedback.
Standard workflow order, tool selection matrix, and composition patterns for qsv CSV data wrangling
Prepare an MCP server and plugin release by bumping versions across all files and updating changelog
Run SQL queries against CSV/TSV/Excel files using Polars SQL engine
Clean a CSV/TSV/Excel file - fix headers, trim whitespace, remove duplicates, validate
Convert between CSV, TSV, Excel, JSONL, Parquet, and other tabular formats
Generate AI-powered Data Dictionary, Description, and Tags for a CSV/TSV/Excel file
| name | review-respond |
| description | Respond to all pending review comments on the current PR — fetch comments, apply fixes, verify accuracy, test, commit, and reply. Use when addressing Copilot reviews, GitHub PR reviews, or any batch of review feedback. |
| argument-hint | [PR number] |
Batch-process all pending review comments on a PR in a single pass.
If a PR number was provided as an argument, use it. Otherwise, detect the current PR:
gh pr view --json number,url,headRefName --jq '.number'
gh api repos/{owner}/{repo}/pulls/{number}/comments --jq '.[] | select(.position != null) | {id, path, line: .original_line, body, user: .user.login}'
Also check for PR review threads with unresolved status:
gh api graphql -f query='{ repository(owner:"{owner}", name:"{repo}") { pullRequest(number:{number}) { reviewThreads(first:100) { nodes { isResolved comments(first:10) { nodes { body author { login } path line } } } } } } }'
copilot or github-actions): Invoke /copilot-review {PR number} to handle these — it fetches, evaluates, applies, and replies automatically.For each unresolved human review comment:
Locate the code with Serena, not bare Read:
mcp__serena__get_symbols_overview on the referenced file to see what's there.mcp__serena__find_symbol (with include_body=true only when you need the body) to jump straight to the function/method/struct the comment is about. Use name_path_pattern like Args/cat_rowskey rather than scanning the whole file.mcp__serena__find_referencing_symbols before changing any signature, return type, or removing a symbol — confirms the blast radius across the codebase.Read when the comment is about plain text (docs, comments, USAGE strings) that has no symbolic structure, or when Serena returns empty.Verify library/API claims with Context7 before agreeing or disagreeing:
byte_headers() consumes the row", "tokio's spawn_blocking returns JoinHandle"), call mcp__context7__resolve-library-id then mcp__context7__query-docs to confirm before acting. Reviewer claims are not always correct; library behavior can change between versions.Apply the fix using Edit (preferred) or mcp__serena__replace_symbol_body for whole-symbol rewrites.
Before committing, verify accuracy of all changes:
Bash with ls/test -f).mcp__serena__find_symbol to confirm it exists at the path you wrote — not Grep. A grep hit on the name is not proof the symbol resolves there.--help output.Run the appropriate test suite based on what was changed:
.rs): cargo test -F all_features.ts): npm test (from .claude/skills/)cargo t {command_name} -F all_featuresgit add <changed files>
git commit -m "address review: <concise summary of fixes>"
For each comment that was addressed, reply via the GitHub API:
gh api repos/{owner}/{repo}/pulls/{number}/comments/{comment_id}/replies -f body="Fixed: <what was changed>"
Output a summary table:
| Metric | Value |
|---|---|
| Comments addressed | N |
| Copilot (via /copilot-review) | N |
| Human reviewer | N |
| Tests | passed/failed/skipped |
| Commit | <hash> |