The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8). * Ingestion points: The script scripts/security_scan.py reads file content from a user-provided directory. * Boundary markers: No explicit delimiters are used to separate scanned content from agent instructions in the output. * Capability inventory: The skill is granted the Bash tool, and the scripts/security_scan.py script performs subprocess calls and has filesystem read access. * Sanitization: The script captures snippets from scanned files without sanitization, which are then included in the scan results.
[COMMAND_EXECUTION]: The script scripts/security_scan.py executes system commands. * Evidence: It utilizes subprocess.run to call npm audit for dependency analysis when a Node.js project is detected.

vulnerability-scanner