auto-swarm

Orchestrate parallel subagent swarms to decompose and distribute complex tasks.

8|Updated Mar 17, 2026
One-click install
npx skills add https://github.com/ShaheerKhawaja/ProductionOS --skill auto-swarm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: auto-swarm
Source: https://github.com/ShaheerKhawaja/ProductionOS/tree/main/skills/auto-swarm
Command: npx skills add https://github.com/ShaheerKhawaja/ProductionOS --skill auto-swarm

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinating complex tasks across multiple subagents is tedious and error-prone, leading to inconsistent results and slower delivery.

Core Features & Use Cases

  • Decomposition & planning: automatically breaks tasks into independent slices and builds a coverage map.
  • Controlled parallelism: runs subagent waves only when beneficial, with configurable depth and swarm size.
  • Convergence & visibility: tracks progress and stops when coverage converges to the desired quality.

Quick Start

Invoke the auto-swarm workflow with a defined task to begin decomposing and distributing work across the swarm.

Frequently Asked Questions about auto-swarm

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

FAQPage Schema
How do I orchestrate parallel subagents for complex software engineering tasks?

To orchestrate parallel subagents, the auto-swarm workflow automatically decomposes complex tasks into independent slices and distributes them across controlled waves. It tracks convergence to ensure desired quality before stopping.

What is subagent swarm orchestration and how does it handle task decomposition?

Subagent swarm orchestration coordinates multiple agents by breaking large tasks into independent slices and building a coverage map. It runs waves of subagents only when beneficial, tracking progress until coverage converges to the desired quality.

Can I configure the swarm size and depth for parallel AI code review workflows?

Yes, you can configure both swarm size and depth for AI code review workflows. The orchestration applies explicit task boundaries and guardrails while running subagent waves only when beneficial for the specified convergence criteria.

What is the best way to decompose large tasks into parallel subagent workflows?

The best way to decompose large tasks is to invoke the auto-swarm workflow with a defined task. It automatically breaks the task into independent slices, builds a coverage map, and distributes work across controlled parallel subagent waves.

Does auto-swarm require any dependencies to run parallel subagent orchestration?

No, auto-swarm requires no external dependencies to run parallel subagent orchestration. It operates as a standalone workflow that manages task decomposition, controlled parallelism, and convergence tracking internally.

When should I not use parallel subagent swarms for task decomposition?

You should avoid using subagent swarms when tasks cannot be broken into independent slices or when controlled parallelism provides no benefit. The workflow explicitly runs subagent waves only when beneficial for convergence and coverage.