rehabilitation-analyzer

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes potentially untrusted data from local rehabilitation files (rehabilitation-tracker.json and daily logs) which serves as an indirect prompt injection surface.
  • Ingestion points: Reads files from the 'data/' directory (SKILL.md).
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the processed data.
  • Capability inventory: Uses file management tools (Read, Grep, Glob, Write, Edit) but lacks network or code execution capabilities.
  • Sanitization: No input sanitization or validation of the ingested JSON content is specified. This finding is considered low risk as there is no exfiltration path or remote execution capability.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 05:32 AM