The Agent Skills Directory

[SAFE]: The skill provides static analysis patterns and dynamic auditing procedures for security assessments. It focuses on identifying misconfigurations in third-party code and infrastructure rather than performing any malicious actions itself.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface as it is designed to ingest and analyze untrusted repository content.
Ingestion points: Analyzes repository files including configuration files, source code, and deployment manifests (as described in the 'Where To Look' section).
Boundary markers: The skill instructs the agent to resolve 'effective configuration' and adhere strictly to the 'engagement scope'.
Capability inventory: Employs shell utilities like ripgrep for file searching and curl for authorized dynamic probes.
Sanitization: Includes a mandatory redaction rule (Procedure E) requiring truncation of discovered secrets and immediate rotation triggers to prevent accidental exposure of sensitive data in reports.
[COMMAND_EXECUTION]: Provides detailed ripgrep (rg) and grep commands intended for the agent to use when auditing local repository files for security vulnerabilities.
[EXTERNAL_DOWNLOADS]: References authoritative security documentation and standards from trusted organizations including OWASP, MITRE CWE, MDN, and official framework documentation (Django, Spring).

aegis-configuration-hardening