security-scan

What problem does it solve?

Manual security reviews are slow, error-prone, and often miss critical vulnerabilities. This Skill automates the detection of common security flaws, ensuring code is safer before it reaches production, saving you from costly incidents and tedious manual checks.

Core Features & Use Cases

• Dynamic Mode Selection: Automatically switches between fast "Basic Mode" for initial reviews and comprehensive "Advanced Mode" for deeper analysis based on revision count.
• Multi-Language Support: Scans Python, JavaScript, Go, Ruby, and Java projects using industry-standard tools like Bandit, Semgrep, npm audit, and SpotBugs.
• Vulnerability Detection: Identifies SQL injection, XSS, hardcoded secrets, insecure dependencies, authentication bypasses, and more.
• Use Case: As a Tech Lead, you can automatically trigger this Skill on every pull request. It will quickly flag high-severity issues in early revisions, and perform a deep dive on later revisions, ensuring no critical vulnerability slips through to production.

Quick Start

To run a basic security scan on the current codebase: export SECURITY_SCAN_MODE=basic bash .claude/skills/security-scan/scan.sh To run an advanced scan: export SECURITY_SCAN_MODE=advanced bash .claude/skills/security-scan/scan.sh