The Agent Skills Directory

[SAFE]: The skill acts as an advisor and auditor for managing other agent capabilities. It does not perform sensitive operations outside of its meta-management scope.
[COMMAND_EXECUTION]: The skill instructions call for the execution of npx @ericrisco/rsc capabilities. This is a vendor-owned utility (belonging to the author ericrisco) used to list installed skills and agent files. Its use is consistent with the skill's primary purpose of environment auditing.
[DATA_EXPOSURE]: The skill records its audit findings in a local file named skill-gaps.jsonl. The instructions include explicit guardrails for the agent to use its own observations and avoid capturing raw user input or sensitive data in these logs, protecting the user's privacy boundary.
[EXTERNAL_DOWNLOADS]: The documentation provides template commands for the user to manually install skills via marketplace plugins or remote tarballs. These are presented as educational examples for the user to follow and do not represent automated or hidden network activities by the agent itself.

skill-scout