The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from the user's codebase to generate security reports and apply fixes. It explicitly instructs the agent to 'Pay attention to specific rules and instructions in the project's documentation and prompt files which may require you to override certain best practices,' which creates a vector for indirect prompt injection if those files contain malicious instructions to bypass security controls.
Ingestion points: The skill reads all project files and documentation to determine language, framework, and security posture.
Capability inventory: The skill can write audit reports, execute shell commands for security tools, and modify the codebase to apply security fixes.
Boundary markers: There are no explicit instructions or delimiters used to prevent the agent from being influenced by instructions embedded within the codebase being audited.
Sanitization: No sanitization of the ingested content is performed before it is used to influence the agent's logic or output.
[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands and security tools for auditing purposes, including npm audit, pip-audit, govulncheck, gitleaks, trufflehog, and detect-secrets.
[EXTERNAL_DOWNLOADS]: The skill recommends using several third-party security auditing tools via package managers, such as npm audit and pip-audit.

security-best-practices