swarming

Coordinate multiple worker agents executing beads with rescue coordination via Agent Mail.

Updated Apr 4, 2026
One-click install
npx skills add https://github.com/anionzo/skill --skill swarming-anionzo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: swarming
Source: https://github.com/anionzo/skill/tree/main/skills/swarming
Command: npx skills add https://github.com/anionzo/skill --skill swarming-anionzo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill provides orchestrated coordination for multiple autonomous worker agents, ensuring they start in a synchronized manner, stay on task via a live bead graph, and recover gracefully from blockers through Agent Mail.

Core Features & Use Cases

  • Phase-based, fault-tolerant swarm orchestration with a dedicated orchestrator and self-routing workers.
  • Live graph monitoring, startup handshakes, and ongoing tend-and-loop coordination for complex multi-bead workflows.
  • Supports file reservations, block handling, and cross-bead handoffs to maximize throughput in large projects.

Quick Start

Launch a swarm of workers to execute the feature beads and report results back through Agent Mail.

Frequently Asked Questions about swarming

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate multiple worker agents for parallel execution?

Coordinate multiple worker agents by decomposing projects into independent beads and using an orchestrator to route workers in parallel. Agent Mail and a live bead graph manage cross-bead handoffs and synchronization to keep execution resilient.

What is fault-tolerant swarm orchestration and when do I need it?

Fault-tolerant swarm orchestration synchronizes autonomous worker agents to execute parallel tasks and recover from blockers. You need it when workflows require cross-bead handoffs, file reservations, and real-time coordination to maintain throughput in large projects.

How do I handle blocked tasks during parallel agent execution?

Handle blocked tasks during parallel agent execution through built-in rescue coordination and escalation protocols. Workers poll inboxes via Agent Mail to receive block handling instructions, ensuring swarms recover gracefully without manual intervention.

Can I use a bead graph workflow for large-scale project decomposition?

Yes, a live bead graph workflow supports large-scale project decomposition by tracking independent beads. It enforces startup handshakes, monitors ongoing tend-and-loop coordination, and manages file reservations to maximize throughput across complex multi-bead workflows.

Does parallel swarm coordination require specific dependencies to run?

Parallel swarm coordination requires no external dependencies to run. It relies entirely on internal mechanisms like Agent Mail for inbox polling, startup handshakes, and bead completion reporting to keep swarms productive and safe.

What is the best way to manage cross-bead handoffs in a multi-agent swarm?

The best way to manage cross-bead handoffs is through Agent Mail and a dedicated orchestrator. This approach enforces startup handshakes and real-time bead graph monitoring, distinguishing it from simpler parallel execution methods by adding robust fault tolerance.