humanize-chinese

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes local Python scripts and a unified CLI tool to perform AI detection and text rewriting tasks. These operations are conducted using the allowed exec tool to process files provided by the user.\n- [PROMPT_INJECTION]: The skill handles user-provided text files, which presents a surface for indirect prompt injection. However, the instructions are specifically designed for linguistic analysis and pattern replacement, minimizing the risk of the agent executing embedded data as instructions.\n
  • Ingestion points: Processes external text files such as manuscripts and papers specified in the command-line examples in SKILL.md.\n
  • Boundary markers: There are no explicit delimiters or boundary markers defined in the prompt instructions to isolate untrusted user data.\n
  • Capability inventory: The skill is configured with access to Read, Write, Edit, and exec tools for its primary text processing functions.\n
  • Sanitization: No explicit sanitization or filtering of the input text is mentioned in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 05:36 AM
Security Audit — agent-trust-hub — humanize-chinese