⚡ MCP Config Generator

MCP Config Generator

Generate the perfect MCP configuration for your AI tool in seconds. Supports Claude Desktop, Cursor, Windsurf, VS Code, Zed, and 100+ more clients.

1Choose your AI client

2Choose installation method

3Install the MCP server

$ pip install aimemory-mcp-server

4Add to Claude Desktop config

Config file location: ~/.claude/claude_desktop_config.json

{
  "mcpServers": {
    "ai-memory": {
      "command": "aimemory-mcp-server",
      "args": []
    }
  }
}

5Restart and test

1Restart Claude Desktop
2Start a new conversation and ask: "Search my memories for [topic]"
3The AI will use the search_memory tool to find relevant conversations

Available MCP Tools

🔍

search_memory

Full-text search across all conversations

📝

add_memory

Save new conversations or memory snippets

🧠

get_context

Retrieve relevant context for the current query

📋

list_memories

List all saved conversations with metadata

📖

get_memory

Get a specific conversation by ID

✏️

update_memory

Update tags, title, or content of a memory

🗑️

delete_memory

Remove a conversation from storage

📊

memory_stats

Get statistics about your memory store

📤

export_memories

Export all memories as JSON

📥

import_memories

Import conversations from a JSON export

🏷️

get_all_tags

List all tags used across conversations

🧹

clear_all

Delete all memories (use with caution)

Frequently Asked Questions

What is MCP (Model Context Protocol)?▼

MCP is an open standard by Anthropic that lets AI assistants connect to external tools and data sources. It enables Claude, Cursor, and 113+ other AI clients to search your conversation history, add memories, and more.

Hosted vs Local — which should I choose?▼

Local (pip/uvx) gives you full offline support and 12 tools. The hosted API requires an internet connection and uses your aimemory.pro session data. Choose local for privacy and full features; choose hosted for zero-setup convenience.

Do I need to upload my conversations first?▼

For the hosted API: yes, upload via aimemory.pro first. For the local server: it has its own SQLite database — you can add memories directly via the AI assistant or import from ChatGPT/Claude exports.

Is this free?▼

Yes! The MCP server is completely free and open-source. Install it, use it, self-host it — no limits, no accounts required for the local version.

Which clients support MCP?▼

Over 113 clients including Claude Desktop, Cursor, Windsurf, VS Code (with Cline/Continue), Zed, Replit, and many more. See the full list at modelcontextprotocol.io.

Ready to give your AI persistent memory?

Upload your ChatGPT, Claude, or DeepSeek conversations and let any MCP-compatible AI assistant search through them.

Upload Conversations Full MCP Docs