Skip to content

For vibecoders

Ship agents that remember, without building memory infra.

You move fast with Cursor and Claude. Myco gives your agents persistent, source-traceable memory through one MCP server, keyless on your own Postgres, so you ship the product, not the plumbing.

Get started in minutes →View on GitHub →

You're rebuilding memory plumbing on every project.

The least fun part, every time
Every new app re-implements the same context wiring. It's the most fragile code you write and the part you least want to own.
Your agent forgets between sessions
You re-explain the same context on every run, and nothing it learns sticks around.
RAG glue you don't want to maintain
A vector store, chunking, session logic. It breaks the moment you ship and look away.
See it in action

Memory your agents get for free.

Wire up one MCP server and your agent remembers, with receipts. No infrastructure to build.

You shipped a feature last week
What did we decide for the onboarding flow, and why?
✓

Myco returns it from durable memory with the source, even in a brand-new session. You never wrote a memory system.

onboarding-notes.mdbrain_why
One line, not a memory system
npm i @mycobrain/mcp-server, point your agent at it, done. Persistent memory without the plumbing.
Keyless and free to run
Full-text, semantic, and graph search all run with no API keys, on your own Postgres. Ship without a bill.
Receipts built in
brain_why shows the source for any answer, so you can trust what your agent tells your users.
Industries

Where Myco is indispensable.

Pick your world. Here is what becomes possible for vibecoders, in the industries where memory you can audit is a requirement, not a nice to have.

FinTech

Regulated

The problem: Regulators now demand a replayable audit trail for every AI decision.

What you can build

Ship a compliance copilot that cites the rule and source behind every answer, without building an audit system.

Why Myco: Every decision traces to its source and rule via brain_why, the program writes the facts (no silent hallucination into the record), and it self-hosts so financial data never leaves your VPC.

Run it. Seed it. Trust it.

01
Run it
Self-host with one docker compose up and npm i @mycobrain/mcp-server. Full-text search, semantic search, and the knowledge graph all run keyless, no API keys required.
02
Seed it
Point Cursor, Claude, or your own agent at Myco, then import your ChatGPT and Claude history, folders, and repos. One deduped, provenance-tracked document per conversation.
03
Trust it
Agents read and write through 11 MCP tools. The program writes the facts, corroboration raises confidence, contradiction supersedes the stale fact, and brain_why traces any answer to its source.

Benchmarked in the open

  • ✓73.6% end-to-end QA on the full 500-question LongMemEval oracle subset, no sampling.
  • ✓recall@5 of 89.2% keyless, 91.6% with the recency reranker, on the full set.
  • ✓Apache-2.0 and free to self-host. No bill to ship.
  • ✓Every number reproduces from a benchmark that ships in the repo.
One MCP server
memory without the plumbing
Keyless
free to run, no API keys
brain_why
receipts on every answer

Ship the product, not the plumbing.

Self-host it free, wire up your agent, and ship something that remembers. Managed cloud is on the waitlist.

Get started in minutes →View on GitHub →
Myco for vibecoders