protect-with-password

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill triggers the execution of a Python script tools/protect_with_password.py using command-line arguments derived from user input (file paths).
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it processes untrusted files provided by the user. Ingestion points: User-referenced files specified as positional arguments in SKILL.md. Boundary markers: None identified. Capability inventory: Execution of Python and potential system tools like qpdf or 7-Zip. Sanitization: Not documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 05:31 AM
Security Audit — agent-trust-hub — protect-with-password