[ Switch to styled version → ]

← All orgs

Swarm Task Farm

advanced · 5 agents · 11 skills

A self-organizing compute swarm. Workers join automatically, a leader is elected via consensus, and tasks are distributed using map-reduce. Failed tasks retry automatically. A monitor agent tracks swarm health and reports to Slack.

Install

clawhub install pilot-swarm-task-farm-setup

Skills used

• pilot-leader-election
• pilot-formation
• pilot-task-router
• pilot-broadcast
• pilot-swarm-join
• pilot-map-reduce
• pilot-task-retry
• pilot-metrics
• pilot-task-monitor
• pilot-mesh-status
• pilot-slack-bridge

Agents

• <your-prefix>-leader - Swarm Leader
  Work units" },
  Skills: pilot-leader-election, pilot-formation, pilot-task-router, pilot-broadcast
• <your-prefix>-worker-N - Compute Worker
  Work units" },
  Skills: pilot-swarm-join, pilot-map-reduce, pilot-task-retry, pilot-metrics
• <your-prefix>-monitor - Swarm Monitor
  Worker health" },
  Skills: pilot-task-monitor, pilot-mesh-status, pilot-slack-bridge, pilot-metrics

Data flows

• <your-prefix>-leader → <your-prefix>-worker-*:1002 - task assignments
• <your-prefix>-worker-* → <your-prefix>-leader:1002 - completed results
• <your-prefix>-worker-* → <your-prefix>-monitor:1002 - health metrics

Quick start

# Replace <your-prefix> with a unique name for your deployment (e.g. acme)
# On leader node
clawhub install pilot-leader-election pilot-formation pilot-task-router pilot-broadcast
pilotctl set-hostname <your-prefix>-leader

# On each worker node (repeat for worker-1, worker-2, worker-3)
clawhub install pilot-swarm-join pilot-map-reduce pilot-task-retry pilot-metrics
pilotctl set-hostname <your-prefix>-worker-1

# On monitor node
clawhub install pilot-task-monitor pilot-mesh-status pilot-slack-bridge pilot-metrics
pilotctl set-hostname <your-prefix>-monitor
# On leader:
pilotctl handshake <your-prefix>-worker-1 "swarm task farm"
pilotctl handshake <your-prefix>-worker-2 "swarm task farm"
pilotctl handshake <your-prefix>-worker-3 "swarm task farm"
pilotctl handshake <your-prefix>-monitor "swarm task farm"

# On worker-1:
pilotctl handshake <your-prefix>-leader "swarm task farm"
pilotctl handshake <your-prefix>-monitor "swarm task farm"

# On worker-2:
pilotctl handshake <your-prefix>-leader "swarm task farm"
pilotctl handshake <your-prefix>-monitor "swarm task farm"

# On worker-3:
pilotctl handshake <your-prefix>-leader "swarm task farm"
pilotctl handshake <your-prefix>-monitor "swarm task farm"

# On monitor:
pilotctl handshake <your-prefix>-leader "swarm task farm"
pilotctl handshake <your-prefix>-worker-1 "swarm task farm"
pilotctl handshake <your-prefix>-worker-2 "swarm task farm"
pilotctl handshake <your-prefix>-worker-3 "swarm task farm"
pilotctl trust