The Agent Skills Directory

[SAFE]: The skill's primary function is to provide the agent with a set of linguistic and structural rules for writing and editing prose. These rules focus on directness, rhythm, and the avoidance of overused AI vocabulary.
[EXTERNAL_DOWNLOADS]: The evaluation metadata in evals/benchmark.json contains a reference to a benchmarking script on the vendor's GitHub repository (github.com/b-open-io/prompts). This is a legitimate reference to the author's own development tools and is not used for runtime execution within the skill.
[SAFE]: The manifest file .clawnet/unsigned-skill.json includes a signature and a hex-encoded opReturnHex payload. This is a standard platform-specific mechanism for skill verification and timestamping, containing a copy of the skill's metadata rather than hidden instructions or executable content.
[DATA_EXFILTRATION]: No network operations or sensitive file access patterns were detected. The skill performs text-to-text transformations within the agent's safe environment.

humanize