Can we have a Test MCP that does code review?

Test MCP for Code Review

Yes — you can build an MCP server that automates code review using linters, type checkers, and static analysis tools.

Code Review MCP Server

pythonCopy
from mcp.server import Server
from mcp.server.stdio import stdio_server
import mcp.types as types
import subprocess, json

app = Server("code-review-mcp")

@app.list_tools()
async def list_tools():
    return [
        types.Tool(
            name="lint_python",
            description="Run ruff linter on a Python file and return issues",
            inputSchema={
                "type": "object",
                "properties": {"file_path": {"type": "string"}},
                "required": ["file_path"]
            }
        ),
        types.Tool(
            name="type_check",
            description="Run mypy type checker on a Python file",
            inputSchema={
                "type": "object",
                "properties": {"file_path": {"type": "string"}},
                "required": ["file_path"]
            }
        ),
        types.Tool(
            name="run_tests",
            description="Run pytest for a specific module",
            inputSchema={
                "type": "object",
                "properties": {"module": {"type": "string"}},
                "required": ["module"]
            }
        )
    ]

@app.call_tool()
async def call_tool(name: str, arguments: dict):
    if name == "lint_python":
        result = subprocess.run(
            ["ruff", "check", arguments["file_path"], "--output-format=json"],
            capture_output=True, text=True
        )
        issues = json.loads(result.stdout) if result.stdout.strip() else []
        msg = f"{len(issues)} issues found" if issues else "Clean — no issues"
        return [types.TextContent(type="text", text=msg)]

    if name == "type_check":
        result = subprocess.run(
            ["mypy", arguments["file_path"]], capture_output=True, text=True
        )
        return [types.TextContent(type="text", text=result.stdout or "Type check passed")]

    if name == "run_tests":
        result = subprocess.run(
            ["pytest", f"tests/test_{arguments['module']}.py", "-v", "--tb=short"],
            capture_output=True, text=True
        )
        return [types.TextContent(type="text", text=result.stdout)]

Register in Claude Config

jsonCopy
{
  "mcpServers": {
    "code-review": {
      "command": "python3",
      "args": ["/path/to/code_review_server.py"]
    }
  }
}

Workflow: Claude + Code Review MCP

textCopy
User: "Review src/auth.py for quality issues"

Claude automatically:
  1. Calls lint_python("src/auth.py") → "3 issues found"
  2. Calls type_check("src/auth.py") → "2 type errors"
  3. Reads file via filesystem MCP
  4. Synthesizes: "Found 5 issues. Here's how to fix each..."

Real Code Review Tools to Wrap

Key Insight

The MCP server handles tool execution — Claude synthesizes results and provides actionable recommendations. This is far more powerful than copying code into a browser chat.