What problem does it solve?
It resolves the common “what should I do now?” dead end by surveying the full Disciplined Agentic Engineering project state and recommending the most valuable next action without changing anything.
Core Features & Use Cases
- Project-wide state survey: Reads in-flight features, consolidation backlog, parked ideas, pending human actions, session-log next tasks, charter/manifest autonomy hints, and open fixes.
- Triage into prioritized decision buckets: Groups work into NEEDS YOUR DECISION, OPEN FIXES, READY TO ADVANCE, READY TO DISPATCH, and COULD START, then selects a single top recommendation.
- Advisory, read-only behavior: Produces a recommendation only—no artifacts and no handoff—so it works as a human re-entry point at any “what now?” moment.
Example use case: You return after a pause and want to immediately know whether to unblock a stalled feature, fix a user-impacting defect, or dispatch safe work to another agent.
Quick Start
Tell the AI “/engineer.next” to get the top recommended next action and the supporting triage buckets.