# neomonic (free edition)

Extended context memory for AI agents, as one small MCP server. It keeps what was said, word for
word, and recalls it when a question needs it. Nothing leaves your machine.

The free edition gives you one context, named `neomonic`. Topics inside it keep things apart.
Unlimited named contexts come with neomonic Pro (neomonic.net).

## Install

One file, no installer, no runtime to install.

**Claude Code** (any platform):

    claude mcp add neomonic -- /path/to/0neo mcp

On Windows, use the full path to `0neo.exe`. Restart Claude Code, then ask it to remember
something. Memory is kept in `~/.neomonic/contexts` (Windows: `C:\ProgramData\neomonic\contexts`);
to keep it somewhere else, add that folder after `mcp`.

**Any other MCP client** (Cursor, Claude Desktop and others):

    { "mcpServers": { "neomonic": { "command": "/path/to/0neo", "args": ["mcp"] } } }

**macOS:** files downloaded by a browser are quarantined, and macOS will say the developer cannot be
verified. Clear it once, in the folder you unpacked:

    xattr -dr com.apple.quarantine .

**Linux:** fully static, runs on any x86-64 Linux. `chmod +x 0neo` if your unpacker dropped the bit.

## Several agents, one memory

Run one server and point every agent at it:

    0neo mcp --http 9966
    claude mcp add --transport http neomonic http://127.0.0.1:9966/mcp

Each agent embeds the memory in its own keyphrases: it registers under its model
(`memory_register`), files every memo under its own headings and cues (`memory_file`), and recalls
through them (`memory_pull`, `memory_answer`). Claude, GPT and Grok agents each build their own way
in over the same memos. No GPU and no outside service.

## Check it

    0neo --version
