danieldoderlein/llm-bus

Overview

An open-source coordination engine for AI agents using the Model Context Protocol. It manages atomic claims, a shared ledger, leases, presence, handoffs, and tasks. You can self-host it or use the hosted service at llm-bus.com.

Best for

Best for
Developers prototyping multi-agent systems needing lightweight coordination

Use cases

• Coordinate multiple AI agents on shared tasks
• Manage resource leases and agent presence in multi-agent systems
• Handle agent handoffs and task delegation

How to use

Install

npm run dev

Tools exposed

• post
• read_posts
• ack
• query_events
• whats_new
• claim
• seed_sequence
• latest_claims
• lease
• release
• who_holds
• task_create
• task_assign
• task_start
• task_block
• task_resolve
• task_ship
• list_tasks
• register
• who_is_active

Tested with

Claude Code, Cursor

Example client config

{\n  "servers": [\n    {\n      "name": "my-llm-bus-server",\n      "url": "http://127.0.0.1:8787/mcp",\n      "auth": "Bearer <token>"\n    }\n  ]\n}

Notes

An open-source coordination engine for AI agents using the Model Context Protocol. It manages atomic claims, a shared ledger, leases, presence, handoffs, and tasks. You can self-host it or use the hosted service at llm-bus.com.

3 stars on GitHub. Last updated 2026-06-22. Licensed AGPL-3.0.

Use cases

• Coordinate multiple AI agents on shared tasks
• Manage resource leases and agent presence in multi-agent systems
• Handle agent handoffs and task delegation

Pros

• Open-source with self-hosting option for control
• Provides structured coordination primitives like ledger and leases
• Built on MCP for integration with AI agents

Cons

• Low stars (3) indicates limited community adoption
• Documentation and maturity may be minimal
• Requires TypeScript knowledge for self-hosting

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.