The Agent Skills Directory

[SAFE]: The skill performs logical analysis on local project data to identify gaps and prioritize features based on revenue impact. It does not utilize network tools, request sensitive credentials, or perform any form of code execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data by reading the ICP definition and the project codebase.
Ingestion points: .agents/memory/icp.md and codebase files (via Step 2).
Boundary markers: None explicitly defined.
Capability inventory: The skill is restricted to generating markdown analysis. It does not possess capabilities to write to the file system (outside of the memory file), execute shell commands, or perform network requests.
Sanitization: Not present.
Evaluation: Because the skill's output is purely informational markdown and it lacks dangerous capabilities, the risk associated with untrusted data ingestion is negligible.

roadmap-analyzer