What problem does it solve? Go codebases often ship with subtle security flaws — SQL injection, weak cryptography, path traversal, leaked secrets — that surface only after a breach. This Skill gives an AI agent a structured security engineering workflow to review, audit, and write Go code that resists these vulnerability classes. ## Core Features & Use Cases - Three operating modes: Review mode for PR security analysis, Audit mode that launches parallel sub-agents across five vulnerability domains with DREAD severity scoring, and Coding mode for writing secure new code. - Comprehensive vulnerability coverage: Injection (SQL, command, XSS, SSRF), cryptography (AES-GCM, Argon2id, TLS), filesystem safety (path traversal, ZipSlip), cookies, secrets management, logging, and memory safety — each with bad/good Go code examples and CWE mappings. - Threat modeling and checklists: STRIDE/DREAD methodology, a full security review checklist, and tooling guidance for gosec, govulncheck, race detection, and fuzz testing. - Use Case: Point your agent at a Go pull request and ask for a security review — it traces data flows beyond the diff, flags a SQL concatenation reachable from user input, scores it with DREAD, and suggests the parameterized-query fix. ## Quick Start Ask your agent to run a security audit of this Go repository using the golang-security skill and report findings by severity.