The Agent Skills Directory

[SAFE]: The skill defines legitimate project management workflows with no evidence of malicious intent, obfuscation, or unauthorized access.\n- [COMMAND_EXECUTION]: The skill performs standard file operations (reading/writing ROADMAP.md and session files) and directory scanning. These actions are within the expected scope of a roadmap tool.\n- [PROMPT_INJECTION]: While the codebase scan feature ingests external content from project files and comments, which is a potential surface for indirect prompt injection, the skill mitigates this by requiring the agent to summarize findings and obtain user confirmation for all changes.

track-roadmap