What problem does it solve? Project roadmaps go stale as libraries, standards, and team discoveries change, but nobody pauses implementation to re-validate the plan. This Skill runs a focused research cycle that grounds findings in the actual codebase, scans the outside world, and mutates native roadmap state so the plan reflects current reality. ## Core Features & Use Cases - Research-driven refresh: Picks a scope (a phase, area, or technology), grounds assumptions in the code with ministr, then researches current best practices and library status with serpapi before proposing changes. - Surgical roadmap mutations: Updates stale descriptions, adds newly discovered backlog chunks, marks obsolete items, records re-prioritization proposals for user review, and logs refresh provenance via roadmap_* tools. - Reasoning trace: Records mandatory think_* steps (open, synthesis, close) so every roadmap change is traceable to the research that motivated it. - Use Case: After a major chunk ships, run a refresh on the next phase to verify the chosen WebSocket library is still maintained, add newly discovered work to the backlog, and surface re-prioritization proposals before more code lands. ## Quick Start Type /roadmap-refresh with an optional topic such as a phase name or technology area to research and update that part of the project roadmap.
