claude-vigil

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a set of instructions for managing project state using provided tools. It does not contain executable code, remote downloads, or obfuscated content.
  • [DATA_EXPOSURE]: The skill mentions managing project files but emphasizes respecting .gitignore to avoid derived or sensitive directories like node_modules or venv.
  • [PROMPT_INJECTION]: The skill involves reading project data which presents an inherent indirect prompt injection surface. However, it includes a verification step (vigil_diff) in its recommended workflow, allowing the user to review changes before a restoration is applied.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 09:49 AM
Security Audit — agent-trust-hub — claude-vigil