Show HN: Scan your AI agents for dangerous capabilities

MakerChecker provides a security layer for AI agents by enforcing deny-by-default tool execution and Ed25519-signed audit trails. It integrates with LangChain and CrewAI to separate agent roles from high-risk skills, preventing agents from approving their own actions. Users can run npx @makerchecker/scan to identify consequential actions like shell commands or data deletion.