The Agent Skills Directory

[COMMAND_EXECUTION]: Utilizes common command-line utilities such as grep and wc to perform static analysis of source code. These operations are intended for identifying insecure coding patterns like unparameterized queries or hardcoded credentials within the developer's repository.
[SAFE]: Provides examples of security anti-patterns and testing payloads (e.g., SQL injection strings). These are provided as reference data for the agent to use when generating regression tests or identifying flaws in user-provided code, rather than as malicious instructions to be executed against the system.
[SAFE]: Incorporates standard security auditing practices by instructing the agent to run npm audit and development tools like eslint and vitest to verify the security and integrity of the application code and its dependencies.

gravitas-security