humanizer

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting untrusted text from users to identify and rewrite AI-generated patterns, which creates a potential vulnerability surface.
  • Ingestion points: The skill reads untrusted data from user messages and local files via the read_file tool as described in SKILL.md.
  • Boundary markers: The instructions lack specific guidance for using delimiters or boundary markers to isolate untrusted input from the agent's instructions.
  • Capability inventory: The skill has the capability to modify the filesystem using patch and write_file tools based on its processing of the untrusted input.
  • Sanitization: There is no documented logic for sanitizing or validating the contents of the processed text before it is presented to the user or written back to the disk.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 07:35 PM
Security Audit — agent-trust-hub — humanizer