The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions define clear ethical boundaries, explicitly directing the agent to refuse requests related to AI-detector evasion or academic dishonesty.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to run a local Python script (renhua_lint.py) for text auditing. The script uses standard libraries for regex-based pattern matching and does not perform network operations or access sensitive system directories.
[SAFE]: The skill processes untrusted user-supplied text as its primary function. While this presents an inherent surface for indirect prompt injection, the risk is minimal given the lack of network access and the specialized nature of the processing. \n
Ingestion points: scripts/renhua_lint.py reads content from user-provided file paths or standard input. \n
Boundary markers: No explicit delimiters are used to wrap the input text, which is typical for a document-editing utility. \n
Capability inventory: Access is restricted to file manipulation tools and local Python execution. \n
Sanitization: Input is processed via regular expressions for auditing purposes without being executed as code.

renhua