The Agent Skills Directory

[SAFE]: The skill provides a structured procedure for authoring agent skills. All operations are local and involve creating directories and files according to a specific template.
[COMMAND_EXECUTION]: The skill instructs the agent to run a local script scripts/validate-metadata.py. This script is included in the package and performs deterministic validation of strings (length, characters, and person perspective) using basic regex and set operations. It does not perform network operations or access sensitive files.
[REMOTE_CODE_EXECUTION]: No remote code execution or external downloads from untrusted sources were detected. The skill uses local assets and scripts exclusively.
[DATA_EXFILTRATION]: There are no network-capable commands or attempts to access sensitive data (e.g., credentials, SSH keys, or environment variables).

skill-creator