# Clone the repository
git clone https://github.com/your-org/securevibes_mcp.git
cd securevibes_mcp

# Install with uv
uv sync

{
  "mcpServers": {
    "securevibes": {
      "command": "uv",
      "args": ["run", "python", "-m", "securevibes_mcp"],
      "cwd": "/path/to/securevibes_mcp"
    }
  }
}

run_assessment → run_threat_modeling → run_code_review → run_dast → generate_report
      ↓                  ↓                   ↓               ↓              ↓
 SECURITY.md    THREAT_MODEL.json   VULNERABILITIES.json  DAST_*.json  scan_results.json
                                                                       scan_report.md

"Run a security assessment on /path/to/my/project"

"Run threat modeling on /path/to/my/project"

"Run code review on /path/to/my/project"

"Run DAST on /path/to/my/project targeting http://localhost:8080"

"Generate a security report for /path/to/my/project"

"What's the scan status for /path/to/my/project"

"Show me the THREAT_MODEL.json for /path/to/my/project"

"Show me critical vulnerabilities for /path/to/my/project"
"Get SQL injection vulnerabilities (CWE-89) in /path/to/my/project"
"List vulnerabilities in the auth module"

User: Run a security assessment on /Users/me/myapp

Claude: I've completed the security assessment. Found 45 files across
Python and JavaScript. Detected Flask and React frameworks.

User: Now do threat modeling

Claude: Identified 12 threats across STRIDE categories:
- 3 critical (authentication, SQL injection, privilege escalation)
- 5 high (XSS, session management, data exposure)
- 4 medium (logging, DoS, input validation)

User: Run code review focusing on the auth module

Claude: Found 5 confirmed vulnerabilities in the auth module:
- VULN-001: Hardcoded credentials in config.py:23
- VULN-002: SQL injection in user_lookup.py:45
...

# Run tests
uv run pytest

# Run with coverage
uv run pytest --cov=securevibes_mcp

# Lint
uv run ruff check .

# Format
uv run ruff format .