The Agent Skills Directory

[SAFE]: The skill is entirely instructional, providing a framework for an agent to decompose tasks and route them to appropriate sub-agents or specialists.
[DATA_EXFILTRATION]: Mentions of 'curl' are limited to examples of verification steps for software testing (e.g., checking an API endpoint) and do not involve exfiltrating sensitive data to external servers.
[PROMPT_INJECTION]: No attempts to bypass safety filters or override agent behavior were found. The skill includes a 'Task' field that ingests external user input for planning, which is a standard functional requirement.

delegation-planner