The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local Python script (scripts/prose_audit.py) to perform mechanical audits of prose. The script is self-contained, does not access the network, and only performs string analysis and statistical calculations on input files provided as arguments.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and rewrite user-provided prose. While this creates a surface for indirect prompt injection, the skill lacks dangerous capabilities that could be exploited via malicious text input. The processing logic is limited to stylistic transformation and metric analysis.
[SAFE]: The skill instructions and associated reference files focus entirely on high-quality prose editing. All external URLs provided in the attribution file are for documentation and source credit to well-known resources and repositories.

antislop-writing