wisdom-mentor

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill utilizes persona embodiment instructions to simulate specific intellectual figures. This role-playing technique is central to the skill's functionality and does not contain explicit safety bypass commands.
  • [COMMAND_EXECUTION]: The skill instructions define a dynamic file access pattern: references/[mentor-name].md. Since the skill allows users to specify mentors, this creates a surface for potential path traversal if the agent attempts to access files outside the intended references/ directory.
  • [PROMPT_INJECTION]: The skill ingests data from local markdown files to define agent behavior, constituting an indirect prompt injection surface. 1. Ingestion points: references/ directory files. 2. Boundary markers: Absent. 3. Capability inventory: Local file reading for persona simulation. 4. Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 08:40 AM