swarm-coordination

Coordinate multi-agent OpenCode swarm workflows with task decomposition and file reservations.

Updated Jan 14, 2026
One-click install
npx skills add https://github.com/primeinc/swarm --skill swarm-coordination-primeinc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: swarm-coordination
Source: https://github.com/primeinc/swarm/tree/main/packages/claude-code-swarm-plugin/skills/swarm-coordination
Command: npx skills add https://github.com/primeinc/swarm --skill swarm-coordination-primeinc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

OpenCode swarm workflows often require coordinating multiple agents to decompose tasks, assign work, and integrate results across files. This skill provides a structured, rules-driven approach to multi-agent coordination, reducing manual orchestration and improving throughput.

Core Features & Use Cases

  • Initialize Swarm Mail orchestration and provide persistent collaboration context.
  • Plan and Decompose tasks into parallelizable subtasks and generate worker assignments.
  • Spawn and Reserve workers with explicit file reservations to prevent conflicts.
  • Review and Handoff check results before finalizing and recording outcomes.
  • Learn and Adapt leverage past learnings via hivemind to improve future tasks.

Quick Start

Use the swarm-coordination skill to plan and coordinate a multi-file task that benefits from parallel agents.

Frequently Asked Questions about swarm-coordination

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

FAQPage Schema
How do I coordinate multiple agents for parallel work in OpenCode?

Multi-agent coordination in OpenCode is handled by decomposing tasks into parallelizable subtasks, spawning workers with explicit file reservations to prevent conflicts, and orchestrating execution through Swarm Mail for persistent collaboration context.

When do I need multi-agent swarm workflows for task decomposition?

Multi-agent swarm workflows are needed for tasks that touch 3 or more files, require parallel work such as frontend, backend, or tests, or benefit from specialized coordinating agents to decompose and assign work across multiple agents.

How does file reservation work when spawning parallel agents?

File reservation during worker spawning assigns explicit file ownership to each agent, preventing write conflicts and ensuring parallel work across frontend, backend, and tests can proceed without overwriting each other's results.

Can I use Swarm Mail to provide persistent collaboration context across agents?

Swarm Mail initializes orchestration and provides persistent collaboration context, allowing multiple spawned agents to coordinate their task execution, share updates, and integrate results across files without losing communication state.

What's the best way to review and handoff results before finalizing multi-agent tasks?

The best approach is to check worker results before finalizing, recording outcomes and leveraging past learnings via hivemind to adapt and improve future task decomposition, worker spawning, and coordination across multi-agent workflows.

Does swarm coordination require MCP lifecycle initialization before spawning workers?

MCP lifecycle initialization is required before task planning and worker spawning, establishing the orchestration layer that enables Swarm Mail communication, file reservations, and post-task review across all coordinated agents.

Related Skills