vers-swarm

Orchestrate parallel agents across branched Vers VMs from a golden state.

2|4|Updated Feb 5, 2026
One-click install
npx skills add https://github.com/hdresearch/pi-vers --skill vers-swarm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: vers-swarm
Source: https://github.com/hdresearch/pi-vers/tree/main/clawdbot/skills/vers-swarm
Command: npx skills add https://github.com/hdresearch/pi-vers --skill vers-swarm

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the need to run multiple agents simultaneously on isolated environments, enabling parallel exploration, distributed problem-solving, and efficient task execution across a swarm.

Core Features & Use Cases

  • Parallel Agent Execution: Spawn and coordinate multiple agents across branched Vers VMs, each starting from an identical golden state.
  • Isolated Environments: Ensures agents operate independently, preventing interference and allowing for diverse approaches to a problem.
  • Use Case: Explore different coding solutions in parallel by branching three VMs, assigning one agent to each VM to test a specific library or implementation strategy, and then comparing the results.

Quick Start

Spawn a swarm of three agents from the golden commit ID 'commit_abc123' to explore different solutions.

Frequently Asked Questions about vers-swarm

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

FAQPage Schema
How do I run parallel agents on isolated environments for distributed problem solving?

To run parallel agents on isolated environments, you spawn an agent swarm across branched Vers VMs derived from a common golden state. Each agent operates independently on its branched VM, enabling parallel exploration and distributed work without interference.

What is agent swarm orchestration across virtual machines?

Agent swarm orchestration across virtual machines is the process of spawning multiple agents simultaneously on branched Vers VMs. It enables multi-path problem solving by having each agent work independently from an identical committed golden state.

Do I need a committed VM state and API key to spawn parallel agents?

Yes, spawning parallel agents requires a VERS_API_KEY and a committed VM state for branching. The committed state serves as the golden state from which multiple isolated branched VMs are derived for simultaneous agent execution.

Can I test different coding solutions in parallel using branched VMs?

Yes, you can test different coding solutions in parallel by branching multiple VMs from a golden commit ID and assigning one agent to each VM. Each agent explores a specific library or implementation strategy independently, allowing you to compare the results.

What's the best way to coordinate multiple agents for multi-path problem solving?

The best way to coordinate multiple agents for multi-path problem solving is orchestrating a swarm across branched Vers VMs. This approach spawns parallel agents on isolated environments derived from a common golden state, ensuring independent and diverse approaches.

Why should I use branched VMs instead of running multiple agents on a single environment?

Branched VMs ensure agents operate in isolated environments, preventing interference and allowing diverse approaches to a problem. Running multiple agents on a single environment risks conflicts, whereas branching from a golden state maintains consistency while enabling parallel work.