human-write
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a Python script (
scan-ai-words.py) intended for local execution to analyze text files. The script processes text using standard regular expressions to identify specific vocabulary patterns. - [DATA_EXPOSURE]: The provided scanning script reads user-specified files. It performs all processing locally, does not include any network communication capabilities, and only outputs information related to the frequency and location of specific dictionary words found within the input.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user text. However, the processing is limited to counting specific word frequencies using a script that correctly escapes input characters (
re.escape), preventing potential regex injection or other data-driven attacks.
Audit Metadata