Static, parser-free security scanner for web/backend source and Apple platform configuration — as a CLI, an MCP server and Claude Skills.
Skill Claude CodeCodex
Use when the user wants found vulnerabilities fixed — "fix these", "how do I close this", every fix request that follows an audit report. Applies the fix and PROVES that the fix actually worked.
Skill Claude CodeCodex
Use while the user is writing a new feature — a login flow, an API endpoint, a file upload, cookie handling, Keychain use, a call to an external service — to settle which secure defaults go in from the start. Prevention, not audit; it comes in before the code is written.
Skill Claude CodeCodex
Use when the user asks for a security analysis or audit of a project, or asks "is this safe?" — a website, a backend API, an iOS/macOS app or a GitHub repository alike. Presents findings with evidence and says plainly which areas were not covered.
Skill Claude CodeCodex
Use to work out what a system needs to be protected against — when starting a new project, before making an architectural decision, or when the question "is this system secure enough?" comes up. Reads no code; derives the assets, the trust boundaries and the attacker model.