Enterprise DNA
M MCP Servers Developer low

NicolasPrimeau/artel

by Various

Self-hosted coordination layer for AI agent fleets: shared memory, tasks, messaging, and session handoffs. Instances mesh together via feeds and mDNS. An autonomous archivist keeps

N

MCP

NicolasPrimeau/artel

Added 1 June 2026

#agent-memory #ai-agents #claude #fastapi #llm #mcp #mcp-server #multi-agent

Overview

Self-hosted coordination layer for AI agent fleets. It provides shared memory, task management, messaging, and session handoffs. Instances mesh via feeds and mDNS, and an autonomous archivist maintains collective knowledge.

Best for

Best for
Developers who want a lightweight, self-hosted orchestration backbone for multi-agent systems

Use cases

  • Orchestrate multiple autonomous agents with shared context
  • Route tasks and messages between distributed agent instances
  • Maintain coherent collective memory across a fleet of agents

Notes

Self-hosted coordination layer for AI agent fleets. It provides shared memory, task management, messaging, and session handoffs. Instances mesh via feeds and mDNS, and an autonomous archivist maintains collective knowledge.

2 stars on GitHub. Last updated 2026-06-01. Licensed MIT.

Use cases

  • Orchestrate multiple autonomous agents with shared context
  • Route tasks and messages between distributed agent instances
  • Maintain coherent collective memory across a fleet of agents

Pros

  • Self-hosted for full data control and privacy
  • Simple mesh discovery using mDNS
  • Built-in archivist automates knowledge cleanup

Cons

  • Very early-stage project with limited adoption (2 stars on GitHub)
  • Requires self-hosting effort and infrastructure management
  • Python-only, limiting integration with non-Python agents

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

Pros

  • Self-hosted for full data control and privacy
  • Simple mesh discovery using mDNS
  • Built-in archivist automates knowledge cleanup

Cons

  • Very early-stage project with limited adoption (2 stars on GitHub)
  • Requires self-hosting effort and infrastructure management
  • Python-only, limiting integration with non-Python agents