The Agent Skills Directory

[SAFE]: The skill provides a structured framework for academic polishing and 'humanizing' text using stylistic rules. It does not contain prompt injections, data exfiltration mechanisms, or unauthorized privilege escalation attempts.
[REMOTE_CODE_EXECUTION]: The skill includes local Python scripts (scripts/ai_detector.py, scripts/text_analyzer.py) for text analysis. These scripts perform local heuristic analysis on provided text and do not download or execute external code.
[EXTERNAL_DOWNLOADS]: The skill lists jieba as a dependency in scripts/requirements.txt. jieba is a well-known, legitimate library for Chinese text segmentation.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided text for polishing. While this constitutes an ingestion surface, the skill lacks dangerous capabilities such as network access or arbitrary system command execution, and it restricts file output to a designated local draft directory (my-thesis/).
[COMMAND_EXECUTION]: The skill instructions suggest running its included Python analysis scripts. This is standard tool invocation for its intended purpose and does not involve executing arbitrary or unsanitized shell commands.

academic-writing