The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes content from the entire codebase. Maliciously crafted source code or documentation files could potentially influence the agent's behavior during the audit process. Ingestion points: Full codebase access via read_file, grep, and codebase_search (SKILL.md, agents/openai.yaml). Boundary markers: None explicitly defined in the prompt instructions to isolate data from instructions. Capability inventory: Shell command execution via bash (agents/openai.yaml). Sanitization: Not explicitly implemented in the provided prompt logic.
[COMMAND_EXECUTION]: The skill utilizes the bash tool to execute security auditing commands such as cargo-audit, npm audit, and pip-audit (SKILL.md, agents/openai.yaml).
[EXTERNAL_DOWNLOADS]: The skill is configured to access well-known external security advisory databases including the National Vulnerability Database (NVD), MITRE CVE, RustSec, Snyk, and GitHub Advisories for cross-referencing vulnerabilities (SKILL.md).

sec-audit