embedmind-mcp 0.1.0

MCP memory server for AI agents: remember / recall / forget over a single local file
docs.rs failed to build embedmind-mcp-0.1.0
Please check the build logs for more information.
See Builds for ideas on how to fix a failed build, or Metadata for how to configure docs.rs builds.
If you believe this is docs.rs' fault, open an issue.

embedmind-mcp

The EmbedMind MCP memory server: gives any MCP-capable agent (Claude Code, Cursor, custom) persistent memory over a single local file — remember, recall, forget — with no server process, no cloud, no Python.

Most users install the all-in-one CLI, which embeds this server as embedmind serve:

cargo install embedmind
claude mcp add embedmind -- embedmind serve --file ~/.embedmind/memory.mind

This crate ships the server as a standalone binary (embedmind-mcp) for integrations that prefer it directly. The memory engine itself lives in embedmind-core.

Tools exposed

Tool What it does
remember Store a memory (text + metadata; embedded and indexed automatically)
recall Semantic search over everything remembered, best match first
forget Delete one memory by id

The protocol is spoken directly over stdio JSON-RPC — no SDK, no async runtime (see ADR 0009).

Status

Pre-v0.1, under active development.

License

MIT. See the repository.