work-humanizer
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied text for humanization. \n
- Ingestion points: Untrusted text provided by the user in Rewrite mode.\n
- Boundary markers: The instructions do not specify delimiters or instructions to ignore embedded commands in the input text.\n
- Capability inventory: The skill performs text generation and suggests executing a local script (check.py) that reads files.\n
- Sanitization: No validation or sanitization of external content is specified before interpolation into the agent's context.\n- [COMMAND_EXECUTION]: The skill includes and encourages the execution of a local Python script
check.pyviapython check.py <file>for verification purposes. Technical analysis of the script confirms it is limited to standard library regex and statistical functions, performing text analysis without network or sensitive file access.
Audit Metadata