humanizer

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to process untrusted text samples to remove AI-isms, creating a vector for instructions to be embedded in data.
  • Ingestion points: The agent ingests data from user messages and by using read_file to access documents and voice samples (SKILL.md).
  • Boundary markers: The instructions do not prescribe specific delimiters or instruction-ignore wrappers for the text being processed.
  • Capability inventory: The skill uses read_file, patch, and write_file to implement its core functionality (SKILL.md).
  • Sanitization: No explicit sanitization or validation steps are defined for the input text before the agent proceeds to patch or overwrite files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:10 AM
Security Audit — agent-trust-hub — humanizer