swarm

Orchestrate parallel software development across isolated git worktree agents.

1|Updated Feb 24, 2026
One-click install
npx skills add https://github.com/MartinNevlaha/stratus-v2 --skill swarm-martinnevlaha
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: swarm
Source: https://github.com/MartinNevlaha/stratus-v2/tree/main/cmd/stratus/skills/swarm
Command: npx skills add https://github.com/MartinNevlaha/stratus-v2 --skill swarm-martinnevlaha

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables the coordination of multiple AI agents working in parallel on different parts of a software development task, each in its own isolated environment, to accelerate feature implementation and bug fixing.

Core Features & Use Cases

  • Parallel Implementation: Spawns isolated workers in git worktrees for truly parallel development.
  • Workflow Orchestration: Manages a multi-phase workflow including exploration, planning, implementation, verification, and learning.
  • Agent Delegation: Intelligently delegates tasks to specialized AI agents based on domain requirements (backend, frontend, database, etc.).
  • Use Case: When a user requests a complex new feature, this Skill can break it down, assign sub-tasks to different specialized agents (e.g., a backend engineer, a frontend developer, a database expert), and manage their parallel execution and integration.

Quick Start

Use the swarm skill to implement the feature described as 'add user authentication with OAuth2'.

Frequently Asked Questions about swarm

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

FAQPage Schema
How do I use multiple AI agents for parallel software development?

Parallel software development with multiple AI agents is orchestrated by spawning isolated workers in git worktrees, dispatching tickets to specialized agents, and monitoring their concurrent progress across exploration, implementation, and verification phases.

Can I implement complex features by delegating sub-tasks to different specialized agents?

You can implement complex features by delegating sub-tasks to specialized AI agents, such as assigning backend, frontend, and database workloads to distinct agents that execute their assigned tickets concurrently in isolated environments.

How does multi-agent orchestration manage the workflow from planning to verification?

Multi-agent orchestration manages a phased workflow spanning exploration, planning, implementation, verification, and learning, coordinating isolated workstreams to ensure concurrent development tasks integrate correctly.

Do I need git worktrees to enable concurrent AI workstreams for feature implementation?

Git worktrees are required to enable concurrent AI workstreams, as this approach spawns isolated workers in separate worktree environments to allow multiple agents to implement features or fix bugs simultaneously without file conflicts.

What is the best way to orchestrate parallel agentic development for bug fixing?

The best way to orchestrate parallel agentic development for bug fixing is to use a multi-agent swarm that breaks down issues, assigns tickets to specialized agents in isolated worktrees, and monitors their concurrent execution and integration.

When should I avoid using a multi-agent swarm for code generation and task orchestration?

You should avoid using a multi-agent swarm for code generation when tasks lack sufficient complexity for parallelization, as the overhead of orchestrating isolated workers and managing phased workflows outweighs benefits for simple, single-step modifications.