From rag-reviewer
Reviews code diffs for performance risks: N+1 queries, repeated work, bad asymptotics, missing batching/caching, blocking I/O, memory growth. Use when asked for a performance review of a PR or diff.
How this skill is triggered — by the user, by Claude, or both
Slash command
/rag-reviewer:performance-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Standalone: ask the user which diff to review if the scope is not clear:
Standalone: ask the user which diff to review if the scope is not clear:
staged — review only the staged diff;unstaged — review only the unstaged diff;main,
master, develop, trunk);Do not pick a scope yourself unless the user already made it clear. If the resulting diff is empty, stop and say there is nothing to review.
Inside /reviewer_review-pr: the orchestrator provides the diffs of all units (path + patch)
— review those. Call the reviewer MCP tools as needed (search_code,
get_related_symbols, read_file, get_definition, find_callers,
get_changed_file_diff) to verify whether a path is truly performance-sensitive.
Look only for performance and efficiency risks in the selected changes. Ignore style, architecture, tests, and general correctness unless they materially affect performance.
Prioritize findings such as:
/reviewer_review-pr use the reviewer MCP tools: read_file,
search_code, find_callers.critical / high: likely severe latency, throughput, or resource regression on
an important path.medium: meaningful inefficiency or scaling risk that should probably be fixed.low: worthwhile optimization or preventive note, not a blocker.Return only actionable findings.
Return ONLY the findings JSON used by the review pipeline, with
"category": "performance":
{"findings": [{
"category": "performance",
"severity": "low|medium|high|critical",
"file": "<path of the reviewed file>",
"line": <line number in the NEW file or null>,
"side": "RIGHT",
"code_quote": "<exact line from the new file>",
"message": "<what is wrong and why it matters>",
"suggestion": "<short advice or null>",
"fix": {"start_line": N, "end_line": M, "replacement": "<new code>"} | null,
"confidence": 0.0
}]}
Standalone runs may additionally render the findings as a readable list after the JSON.
If a finding cannot be tied to a specific line, use the closest changed line and
explain the scope in message.
If there are no meaningful performance findings, return {"findings": []} and say so.
Write message and suggestion in the output language given by the orchestrator
(standalone: the user's language).
npx claudepluginhub mimfort/rag_for_gitReviews git diffs for reuse opportunities, quality issues, and inefficiencies using three parallel agents, then fixes issues. Triggers on 'simplify', 'clean up', or after code changes.
Reviews pull requests, branch diffs, and local working-tree diffs for correctness, security, concurrency, performance, and code quality issues. Returns structured JSON findings.
Reviews code changes, PRs, and diffs for security vulnerabilities, performance issues, correctness bugs, and maintainability problems with tables for issues and suggestions.