humanizer-paper
Pass
Audited by Gen Agent Trust Hub on Jul 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a stylistic polishing tool for academic drafts. All included components (markdown rules and a Python linter script) are designed for local text processing without external network dependencies or sensitive data access.
- [COMMAND_EXECUTION]: The skill uses a limited
Bash(python *)tool to run its own bundled mechanical linter (scripts/polish_lint.py). The script uses only standard Python libraries (re,statistics,json, etc.) to analyze text cadence and surface markers. It does not perform any dangerous system operations or execute arbitrary code. - [DATA_EXFILTRATION]: No network operations or external data transmissions were detected. The Python script reads from stdin or local files and outputs analysis results directly to the agent's console.
- [PROMPT_INJECTION]: The skill contains explicit '诚信边界' (integrity boundary) instructions that direct the agent to refuse requests for bypassing AI detectors (e.g., Turnitin or 知网 AIGC) and instead refocus the user on improving originality and academic rigour.
- [EXTERNAL_DOWNLOADS]: The skill does not download external code or dependencies. It relies on internal reference files and a pre-bundled Python script.
Audit Metadata