code-review

name	code-review
description	Structured code quality assessment with severity ratings
tags	["quality","review"]
tools_required	["read_file","list_directory"]
parameters	[{"name":"target","type":"string","description":"File or directory to review","required":true},{"name":"focus","type":"string","description":"Review focus area (security, performance, readability, all)","default":"all"}]
output_format	json

Code Review

Perform a structured code quality review of a file or directory, producing actionable findings with severity ratings.

Steps

1. Discover files: If {target} is a directory, use list_directory to enumerate source files (skip binary, vendored, and generated files). If it is a single file, proceed directly.

2. Read and analyze: For each file, use read_file to load the contents. Analyze based on the focus area ({focus}):

   • security: Look for injection risks, hardcoded secrets, unsafe deserialization, missing input validation, path traversal, SSRF, unescaped output.
   • performance: Identify N+1 queries, unbounded loops, missing caching opportunities, excessive allocations, blocking calls in async code.
   • readability: Check naming conventions, function length, dead code, missing docstrings, inconsistent style, overly complex conditionals.
   • all: Apply all of the above categories.

3. Rate each finding: Assign a severity:

   • critical -- security vulnerability or data loss risk, must fix before merge
   • high -- bug or significant performance issue
   • medium -- code smell or maintainability concern
   • low -- style nit or minor improvement suggestion

4. Produce findings: For each issue, record the file path, line number(s) or function name, category, severity, a one-line summary, and a suggested fix.

5. Summarize: Return a JSON object:

   {
     "target": "{target}",
     "focus": "{focus}",
     "files_reviewed": <int>,
     "findings": [
       {
         "file": "...",
         "line": "...",
         "category": "security|performance|readability",
         "severity": "critical|high|medium|low",
         "summary": "...",
         "suggestion": "..."
       }
     ],
     "summary": "Brief overall assessment"
   }
   

Guidelines

• Be specific: cite line numbers and function names, not vague observations.
• Limit findings to genuinely actionable items -- skip style preferences that are already consistent within the codebase.
• If no issues are found, return an empty findings array with a positive summary.