The Agent Skills Directory

[SAFE]: The skill defines a set of instructional guidelines and a scoring system for an AI agent to categorize tasks (e.g., 'quick', 'standard', 'thorough'). This logic is purely descriptive and used for internal task routing.
[NO_CODE]: There are no scripts, executables, or shell commands contained within the skill. It describes integration with other tools (like cavekit-router.cjs) but does not execute them directly.
[SAFE]: The skill does not perform any network operations or access sensitive credentials. It mentions security-sensitive topics only as criteria for increasing the complexity score of a task.

complexity-detection