The Agent Skills Directory

[SAFE]: The skill performs its primary function of project planning and auditing using local file access and user interaction. It does not utilize network access or execute external code.
[NO_CODE]: No scripts, executables, or binary files are included in the skill package. The skill consists entirely of markdown instructions and documentation.
[SAFE]: The skill incorporates human-in-the-loop verification for its codebase scanning features. During the 'Generate' and 'Audit' phases, the agent is instructed to summarize findings and obtain explicit user confirmation before modifying the ROADMAP.md file or creating session work plans.

track-roadmap