MCP Server · npm package

AI-Powered Automated Code Review

An MCP server that gives AI agents 8 tools to automate your code review pipeline. Security scanning, style checking, complexity analysis, dependency auditing, test coverage, and pull request review — all callable programmatically.

Start Reviewing — $25/mo View Pricing

$25/mo. Catch bugs before they ship.

8 MCP tools npm package Works with Claude, Cursor, any MCP client TypeScript
GitHub npm npm

The Problem

Manual code review does not scale

Senior engineers spend 6+ hours per week reviewing pull requests. Reviews become bottlenecks, security issues slip through, and code quality is inconsistent across the team. The bigger your team grows, the worse it gets.

6+ hrs
Per week senior engineers spend on code review
40%
Of security vulnerabilities missed in manual review
3 days
Average PR wait time before first review

The Solution

8 MCP tools that automate your code review pipeline

TRUSS Code Review is an MCP server that AI agents call to review code automatically. No waiting for reviewers. No missed security issues. Your agents scan, analyze, and review — programmatically.

scan_security

Security Scanning

Detect SQL injection, XSS, hardcoded secrets, insecure dependencies, and OWASP Top 10 vulnerabilities. Pattern matching plus semantic analysis for fewer false positives.

check_style

Style Checking

Enforce coding standards across your team. Check naming conventions, formatting, import ordering, and project-specific rules. Configurable per repository.

analyze_complexity

Complexity Analysis

Measure cyclomatic complexity, cognitive complexity, and nesting depth. Flag functions that are too complex, too long, or have too many parameters.

audit_dependencies

Dependency Auditing

Scan package manifests for known vulnerabilities, outdated versions, license conflicts, and typosquat risks. Check npm, pip, go modules, and more.

analyze_coverage

Test Coverage Analysis

Identify untested code paths, missing edge cases, and functions without test coverage. Suggest specific tests to write based on code changes in the PR.

review_pr

Pull Request Review

Full PR review: read the diff, understand context, check for bugs, suggest improvements, and produce a structured review with inline comments and summary.

detect_patterns

Anti-Pattern Detection

Flag common anti-patterns: God objects, deep nesting, magic numbers, dead code, circular dependencies, and language-specific code smells.

generate_review_report

Review Reporting

Generate structured review reports with severity-ranked findings, fix suggestions, and metrics. Track code quality trends across your team over time.

How It Works

From install to automated code review in 60 seconds

TRUSS Code Review is an MCP server. Install it via npm, add it to your agent's MCP config, and your AI agents can call all 8 tools immediately.

  1. 01

    Install the package

    One command. No build step. Works with any MCP-compatible client — Claude Desktop, Cursor, VS Code, or your own agent.

  2. 02

    Configure your review rules

    Set your API key and optional configuration: language-specific rules, severity thresholds, custom patterns to check, and team style guides.

  3. 03

    Agents review code programmatically

    Your AI agent discovers the 8 review tools and calls them as needed. Scan for security issues, check style, analyze complexity, audit dependencies, and generate review reports — all in a single agent conversation.

# Install npx truss-code-review-mcp # Or add to Claude Desktop MCP config { "mcpServers": { "code-review": { "command": "npx", "args": ["truss-code-review-mcp"], "env": { "TRUSS_API_KEY": "your-api-key" } } } }

Pricing

Better code review than your most senior engineer

At $25/month, TRUSS Code Review costs less than 30 minutes of a senior engineer's time — and it never gets tired, never misses a pattern, and reviews every PR in seconds.

Code Review
$25/mo

Everything you need to automate code review with AI agents.

  • All 8 MCP tools
  • Unlimited reviews
  • Security scanning
  • Complexity analysis
  • Dependency auditing
  • Test coverage analysis
  • Anti-pattern detection
  • Review reporting
  • Email support
Get Code Review

Tool Reference

All 8 code review tools

  1. 01 scan_security — Detect vulnerabilities: SQL injection, XSS, hardcoded secrets, OWASP Top 10
  2. 02 check_style — Enforce coding standards, naming conventions, and formatting rules
  3. 03 analyze_complexity — Measure cyclomatic and cognitive complexity, flag overly complex code
  4. 04 audit_dependencies — Scan for vulnerable, outdated, and license-conflicting dependencies
  5. 05 analyze_coverage — Identify untested code paths and suggest specific tests to write
  6. 06 review_pr — Full pull request review with inline comments and structured summary
  7. 07 detect_patterns — Flag anti-patterns, code smells, dead code, and circular dependencies
  8. 08 generate_review_report — Severity-ranked findings, fix suggestions, and quality trend metrics

Stop shipping bugs to production

Automate your code review pipeline with AI agents. One MCP server, 8 tools, $25/month. Every PR reviewed in seconds.

Start Reviewing — $25/mo

$25/mo. Cancel anytime.

View on GitHub npm View on npm