Analyzing Security

Scans your codebase for security vulnerabilities using pattern matching across SQL injection, XSS, path traversal, hardcoded secrets, and other OWASP risks. Automatically runs on new modules, security-related changes, and post-refactor. Critical and high severity findings (injection flaws, exposed credentials) block delivery, while medium and low can be accepted with documented rationale in DESIGN.md. The implementation is surprisingly practical about false positives, offering exclude patterns and context-based severity downgrading. For example, MD5 for hash bucketing or non-crypto random for jitter get a pass. Ships with a Node script that outputs human or JSON formats for CI integration. Chains nicely with other skills like securing-systems for remediation guidance.

Install to Claude Code

npx -y skills add telagod/code-abyss --skill analyzing-security --agent claude-code