Use cases
AI agent memory that follows the user, not one chat window
Use XMemo when an AI agent needs durable context across GitHub Copilot, Claude, ChatGPT, Gemini CLI, Cursor, Codex, and any MCP-compatible client.
Keywords: AI agent memory, agent memory layer, MCP memory, Copilot memory, Claude memory, ChatGPT memory
What XMemo stores
Project facts, preferences, decisions, follow-ups, and reusable context are saved as user-owned memories that can be recalled by authorized agents.
- Cross-client context for IDEs, CLIs, and chat surfaces
- Attribution by agent, device, and client instance
- OAuth or scoped token access without exposing private memory publicly
Why agents need it
Most agents lose context when a session ends or keep memory locked to one product. XMemo provides a separate memory control plane that authorized tools can share.
- Persistent memory across sessions
- A single MCP endpoint for many clients
- Recall that can be grounded without retraining on private data
Frequently asked questions
What is AI agent memory?
AI agent memory is durable user context that an authorized assistant can save and recall across sessions, tools, and devices.
Does XMemo replace a model's built-in memory?
No. XMemo is an external memory layer that works alongside built-in memories and lets the user control cross-client context through MCP.
Which AI agents can use XMemo?
Any client that supports MCP or the XMemo API can use it, including Copilot-style CLIs, Claude, ChatGPT, Gemini CLI, Cursor, and custom agents.