The Agent Skills Directory

[DYNAMIC_EXECUTION]: The secret_scanner.py and threat_modeler.py scripts use __import__('datetime') for report timestamps. This is a standard Python pattern for lazy imports and is safe as the target is a static string for a core library.
[INDIRECT_PROMPT_INJECTION]: The skill ingests file contents during secret scanning and user strings for threat modeling. 1. Ingestion points: scripts/secret_scanner.py reads project files; scripts/threat_modeler.py accepts --component strings. 2. Boundary markers: None explicitly defined, but usage is restricted to structured scanning workflows. 3. Capability inventory: The skill can read local files and write reports; it executes internal analysis scripts via command line. 4. Sanitization: Findings in the secret scanner are masked to avoid unintentional exposure of credentials in standard output.

senior-security