io.github.helinakdogan/agent-magnet
Behavioral memory layer for AI tools. Learns preferences and context across Claude, Cursor, Codex.
What can you do with io.github.helinakdogan/agent-magnet?
- Learns automatically as you work — no "remember this" required
- Shares one memory across Claude, Cursor, and Codex
- Lets you forget or mark things done — memory doesn't just pile up
Install
Add this to your client's MCP configuration (Claude Desktop: claude_desktop_config.json; Cursor: .cursor/mcp.json; VS Code and others use the same "mcpServers" shape).
agent-magnetclaude mcp add agent-magnet -e 'MAGNET_USER_ID=<MAGNET_USER_ID>' -e 'MAGNET_OPENAI_KEY=<MAGNET_OPENAI_KEY>' -e 'MAGNET_LOCAL_MODE=<MAGNET_LOCAL_MODE>' -e 'MAGNET_REDIS_URL=<MAGNET_REDIS_URL>' -e 'MAGNET_PROJECT_ID=<MAGNET_PROJECT_ID>' -- uvx agent-magnet==0.1.11
{
"mcpServers": {
"agent-magnet": {
"command": "uvx",
"args": [
"agent-magnet==0.1.11"
],
"env": {
"MAGNET_USER_ID": "<MAGNET_USER_ID>",
"MAGNET_OPENAI_KEY": "<MAGNET_OPENAI_KEY>",
"MAGNET_LOCAL_MODE": "<MAGNET_LOCAL_MODE>",
"MAGNET_REDIS_URL": "<MAGNET_REDIS_URL>",
"MAGNET_PROJECT_ID": "<MAGNET_PROJECT_ID>"
}
}
}
}
Environment variables
MAGNET_USER_IDrequired Your identifier used as the memory key across all toolsMAGNET_OPENAI_KEYOpenAI API key for the reflector and classifier modelsMAGNET_LOCAL_MODESet to 1 to use local SQLite storage — no Redis or external services neededMAGNET_REDIS_URLRedis connection URL for persistent shared storage (overrides local mode)MAGNET_PROJECT_IDNamespace for memory isolation across projects (defaults to 'default')
Documentation
From the project's README on GitHub, refreshed weekly. Links open on GitHub.
Your AI forgets your project every new session. This fixes that.
What it does
- Learns automatically as you work — no "remember this" required
- Shares one memory across Claude, Cursor, and Codex
- Lets you forget or mark things done — memory doesn't just pile up
Quick start
pip install agent-magnet
agent-magnet init
That's the whole install. No Redis, no keys, local by default.
What it remembers
Decisions, conventions, watch-outs, failed attempts, goals, preferences, and actions taken — one memory, sorted automatically.
Team memory
Shared memory for a team — one person's decision becomes everyone's context. Requires a paid Agent Magnet key from agentmagnet.app.
Works everywhere
Claude, Cursor, Codex — same memory, local by default. agent-magnet init detects and configures Claude Code, Claude Desktop, and Cursor for you.
Manual MCP config (Codex, or any other MCP client)
{
"mcpServers": {
"agent-magnet": {
"command": "agent-magnet-mcp",
"env": { "MAGNET_USER_ID": "your_name" }
}
}
}
MIT licensed. Contribute — and a ⭐ if this saved your context window.
Frequently asked questions
What is io.github.helinakdogan/agent-magnet?
Behavioral memory layer for AI tools. Learns preferences and context across Claude, Cursor, Codex.
How do I install io.github.helinakdogan/agent-magnet?
It runs on your machine. In Claude Code: claude mcp add agent-magnet -e 'MAGNET_USER_ID=<MAGNET_USER_ID>' -e 'MAGNET_OPENAI_KEY=<MAGNET_OPENAI_KEY>' -e 'MAGNET_LOCAL_MODE=<MAGNET_LOCAL_MODE>' -e 'MAGNET_REDIS_URL=<MAGNET_REDIS_URL>' -e 'MAGNET_PROJECT_ID=<MAGNET_PROJECT_ID>' -- uvx agent-magnet==0.1.11. For other clients, paste the JSON configuration from this page into their MCP settings.
Does io.github.helinakdogan/agent-magnet need an API key?
Yes. It expects MAGNET_USER_ID, which you set in its configuration.
Is io.github.helinakdogan/agent-magnet open source?
Yes. Its code is on GitHub under the MIT license.
Who maintains io.github.helinakdogan/agent-magnet?
It is published by helinakdogan. AI.tn lists it but does not maintain or endorse it.
From the official MCP Registry. Listing is not an endorsement: review a server's code and permissions before giving it access to your data.