Log inSubmit server
Ruv Swarm logo

Ruv Swarm

Runs ephemeral neural-agent swarms through an MCP-enabled CLI with Rust and WebAssembly components.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Ruv Swarm

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "ruv-swarm": {
      "command": "npx",
      "args": [
        "-y",
        "ruv-swarm@latest"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

We couldn’t confirm it starts correctly

We ran the install command but it didn't respond within our test window — this can mean a slow first-time install rather than a real problem.

npx -y ruv-swarm@latest

No response to initialize.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked 1mo ago.

Overview

The ruv-swarm MCP server orchestrates ephemeral neural networks for task-specific problem solving through an MCP and CLI interface. It creates specialized networks on demand, executes them with CPU-native WebAssembly, and removes them after completion. The project also exposes swarm concepts such as configurable topologies, cognitive patterns, and adaptive learning. Reach for it when an agent workflow needs local, distributed neural orchestration rather than a direct model API.

Use cases

  • Orchestrate specialized neural agents for software tasks
  • Run CPU-based swarm workloads without CUDA
  • Experiment with mesh, ring, or hierarchical agent topologies
  • Initialize Claude Code with the project’s MCP integration

Key features

  • Ephemeral task-specific neural networks
  • MCP and CLI integration
  • WebAssembly execution
  • Five documented swarm topologies
  • Multiple neural architectures
  • Adaptive learning and optimization

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked Today

Recent health check succeeded.

8/8 checks healthy over the last 44d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 6d ago via OSV.dev · ruv-swarm@latest (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars387Stargazers on the source repository.
npm downloads48kPackage downloads in the last 30 days.
Last commit2mo agoMost recent push to the default branch.
Install checkInconclusiveDidn't respond in our test window — often a slow first install.
Directory activity5 viewsConfig copies, upvotes, and views on AllMCPs.

How Ruv Swarm works

What ruv-swarm MCP server does

The ruv-swarm MCP server is the orchestration layer in the larger ruv-FANN repository. It is designed to create short-lived neural networks for specific tasks, coordinate them as a swarm, and dissolve them after the work is complete. The repository describes this as an alternative to simply calling a model API: intelligence is instantiated for a task and then discarded.

The surrounding framework includes ruv-FANN, a Rust implementation of the FANN neural-network library, and Neuro-Divergent, a forecasting project with models including LSTM, N-BEATS, and Transformers. The MCP server sits above these components in the documented architecture and is presented as an integration point for Claude Code and other applications.

How it works

A task begins with on-demand network creation. Each network is specialized for the requested problem, executes using CPU-native WebAssembly, and is removed when its task finishes. The documented architecture places the MCP/CLI layer above the forecasting models and the Rust neural-network engine, with a WebAssembly runtime supporting browser, edge, server, and embedded environments.

The ruv-swarm MCP server describes five available swarm topologies: mesh, ring, hierarchical, star, and custom. It also lists seven cognitive patterns, including convergent, divergent, lateral, and systems thinking.

Setup and configuration

The README gives three installation paths:

  • Run npx ruv-swarm@latest init --claude without a separate installation.
  • Install the package globally with npm install -g ruv-swarm.
  • Install the Rust CLI with cargo install ruv-swarm-cli.

The init --claude command is the documented route for setting up the Claude-oriented MCP integration. No API key, token, or required environment variable is documented here.

Tools and capabilities

Documented capabilities include:

  • Creating ephemeral, task-specific neural networks.
  • Coordinating distributed agents through multiple swarm topologies.
  • Running neural workloads with CPU-native WebAssembly rather than requiring CUDA or a GPU.
  • Using neural architectures ranging from multilayer perceptrons to Transformers.
  • Supporting Claude Code through MCP integration.
  • Generating contribution assistance for code analysis, reviews, tests, documentation, and pull requests through the project’s GitHub swarm workflow.

Treat those figures as project-reported results rather than independently verified guarantees.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Ruv Swarm

Run `npx ruv-swarm@latest init --claude`, or use the documented npm and Cargo installation paths.

The project describes its execution as CPU-native and says no CUDA or GPU is required.

Related MCP servers

All in Developer Tools All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewRuv Swarm AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/ruv-swarm?style=directory)](https://allmcps.com/mcp/ruv-swarm)
HTML Embed
<a href="https://allmcps.com/mcp/ruv-swarm"><img src="https://allmcps.com/api/badge/ruv-swarm?style=directory" alt="Ruv Swarm on AllMCPs" /></a>