nutritional-specialist

Pass

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: SAFE
Full Analysis
  • DATA_EXPOSURE (SAFE): The preferences_manager.py script manages data in ~/.claude/nutritional_preferences.json. This is standard local storage for application-specific data (nutritional goals, allergies) and does not involve unauthorized access to system secrets or personal credentials.
  • REMOTE_CODE_EXECUTION (SAFE): No external downloads or dynamic execution patterns (like eval or piped curl-to-bash) were detected.
  • PROMPT_INJECTION (LOW): The skill has a surface for indirect prompt injection because it processes user-supplied text (notes, preferences). However, no malicious instructions are present in the code.
  • Ingestion points: preferences_manager.py loads data from nutritional_preferences.json.
  • Boundary markers: Absent in the management script.
  • Capability inventory: Local file read/write; no network or subprocess capabilities.
  • Sanitization: The script performs standard JSON parsing but does not sanitize the content of strings, relying on the calling agent for safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 18, 2026, 01:02 AM