The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted source code and configuration files to identify vulnerabilities. This process inherently exposes the agent to an indirect prompt injection surface, where malicious instructions embedded in the analyzed code could attempt to influence the agent's behavior.
[COMMAND_EXECUTION]: The skill workflow involves checking for and potentially executing local security scanning tools such as semgrep, bandit, gosec, brakeman, trivy, and gitleaks. It uses the which command to verify tool availability before use, which is a standard environment check for automated auditing tasks.

pasta-vulns