By name: ntm-next-bead description: Re-enter the worker loop by checking coordination state, picking the next actionable task, and continuing without waiting for operator intervention version: 2.2.0 author: Daniel Fischer category: automation tags: ["ntm", "multi-agent", "swarm", "beads", "workflow"]
Continue to Next Task
Use this when you are already oriented and need to get moving again.
If you are returning from compaction or a long idle period and have not reconstructed state yet, run /ntm-reread-agents first.
Re-entry Rules
- You are still the same fungible worker unless the project explicitly changed your role.
- Reuse the same identity lease; do not create a new one just because context was lost.
- If you already own active work, continue it instead of picking something new.
- If mail or the operator directly assigns you a task, follow that instead of generic triage.
Re-entry Loop
- Check inbox and active threads.
- Confirm whether you already own active work.
- If not, select the next actionable task using the project task flow from
AGENTS.md. - Claim it.
- Reserve the required files or work surface.
- Announce start.
- Implement, polling inbox at natural checkpoints.
- Self-review and run the required quality gates.
- Close or update the task, sync task metadata, release reservations, and follow the project's commit handoff rules.
- Check inbox again.
- Continue the worker loop.
If Nothing Is Actionable
- Run
/ntm-review-others, or - Run
/ntm-unstallif the queue appears blocked or abandoned, - Then re-check the queue.
Do not stop just to summarize the project to the user unless explicitly asked or unless no actionable work remains.