提交

Orbit Shared Memory For Agents

@Fraima

Orbit is an MCP-connected operations layer for Claude and other AI Agents. Connect once and every session — chats, Claude Cowork/Code/ChatGPT Codex and other agents automatically log outputs, reports and decisions to a shared dashboard. Agents read from and write to Orbit's knowledge layer, so context persists across sessions and tools without re-explaining. Includes live agent visibility, trust-weighted memory, broadcast to all agents, and agent creation. Free to start.
概览

Orbit is an MCP-connected operations layer for Claude and other AI tools. Connect once and every session, Claude chats, Claude Code, scheduled agents, automatically logs outputs, files, and decisions to a shared dashboard. What it does:

Live visibility: see every agent and session in real time, one dashboard Shared memory: agents read from and write to a trust-weighted knowledge layer automatically, no manual curation Decision layer: make a decision in Orbit, every connected agent reads it before acting Broadcast: send one instruction to all agents instantly Agent creation: spin up scheduled agents directly from the dashboard AI org chart: visual map of every agent, every connection, every data flow

Who it's for: solo founders, solopreneurs, freelancers, and vibe coders running their business on Claude and AI tools. If you have more than one agent or session running and you've ever lost track of what any of them did — this was built for you. Setup: paste one MCP command into Claude. Two minutes. No webhooks, no OAuth. Free to start. Paid tiers for Insights, State, and Orbit Agents.

服务器配置

{
  "mcpServers": {
    "orbit": {
      "type": "http",
      "url": "https://orbitagents.xyz/api/public/mcp/<YOUR_API_TOKEN>"
    }
  }
}
© 2025 MCP.so. All rights reserved.

Build with ShipAny.