SilentSwarm CLI Guide for LLMs

This document is a static, repository-tracked variant of the runtime-generated output from:

swarm llm-doc

For environment-specific output (for example current leader URL), generate fresh docs via:

swarm llm-doc --output docs/llm_swarm_cli.generated.md

Core Commands

• swarm info
• swarm nodes
• swarm devices
• swarm fellows
• swarm release --fellow-id N | --all
• swarm set-speed --fellow-id N --tx-mbps X --rx-mbps Y
• swarm update [--fellow-id N ...]
• swarm train --config PATH
• swarm sweep --config PATH
• swarm jobs
• swarm stop JOB_ID
• swarm watch JOB_ID

Typical Agent Workflow

1. Inspect topology with nodes/devices/fellows.
2. Normalize fellow set and network limits.
3. Submit one train run or a full sweep queue.
4. Observe with jobs/watch until terminal state.
5. Stop and retry with changed parameters if needed.

API Mappings

• nodes -> GET /api/v1/cluster/nodes
• devices -> GET /api/v1/cluster/devices
• fellows -> GET /api/v1/workers
• release single -> DELETE /api/v1/workers/{fellow_id}
• release batch/all -> POST /api/v1/workers/release
• set-speed -> POST /api/v1/workers/{fellow_id}/network-limit
• update -> POST /api/v1/workers/update
• train/sweep submit -> POST /api/v1/jobs
• jobs -> GET /api/v1/jobs
• stop -> DELETE /api/v1/jobs/{job_id}
• watch -> GET /api/v1/jobs/{job_id}/metrics/stream