scan-vulnerabilities
Detect security vulnerabilities in code and dependencies. Use when auditing security.
When & Why to Use This Skill
This Claude skill automates the detection of security vulnerabilities within source code and third-party dependencies. It streamlines security auditing by identifying unsafe coding patterns, known CVEs, and potential exploits using industry-standard tools like Bandit, Safety, and Semgrep, ensuring robust application security and proactive risk mitigation throughout the software development lifecycle.
Use Cases
- Pre-production Security Audits: Systematically scanning codebases for SQL injection, hardcoded secrets, and unsafe execution patterns before any production release.
- Dependency Vulnerability Management: Identifying and mitigating risks from outdated or compromised third-party libraries by checking against known CVE databases.
- DevSecOps Integration: Incorporating automated security scans into CI/CD pipelines to catch vulnerabilities early in the development cycle and prevent regressions.
- Compliance and Risk Assessment: Generating detailed security reports including severity levels and CVSS scores to meet industry standards and regulatory requirements.
| name | scan-vulnerabilities |
|---|---|
| description | "Detect security vulnerabilities in code and dependencies. Use when auditing security." |
| mcp_fallback | none |
| category | security |
| tier | 2 |
| user-invocable | false |
Scan Vulnerabilities
Systematically scan code for security vulnerabilities including unsafe patterns, known CVEs, and potential exploits.
When to Use
- Regular security audits
- Before releasing code to production
- When updating dependencies
- In CI/CD security checks
Quick Reference
# Python security scanning
pip install bandit safety
# Scan code for security issues
bandit -r . -ll
# Check for known vulnerabilities in dependencies
safety check
# Advanced: SAST scanning
python3 -m pip install semgrep
semgrep --config=p/security-audit --json .
Workflow
- Scan code for issues: Identify unsafe patterns (SQL injection, exec, hardcoded secrets)
- Check dependencies: Scan for known vulnerabilities (CVEs)
- Review findings: Analyze severity and exploitability
- Prioritize fixes: Address critical/high severity issues first
- Document fixes: Record how vulnerabilities were resolved
Output Format
Security scan report:
- Vulnerability type (SQL injection, hardcoded secret, etc.)
- Location (file, line number)
- Severity (critical/high/medium/low)
- CVSS score (if applicable)
- Vulnerable dependency version (if applicable)
- Recommended fix
- Fixed version (if dependency)
References
- See CLAUDE.md > Security standards for security guidelines
- See
quality-security-scanskill for automated CI scanning - OWASP Top 10 for common vulnerability categories