One Brain

io.github.Prophet-gitPublished by GitHub account Prophet-gitv0.1.666Updated Oct 25, 2026

Shared memory for a team in Claude Code: what one person records, everyone has.

The publisher's own server: BrainDog

VerifiedStreamable HTTPWeb executableKnowledge & MemoryCommunication & Collaboration

Overview

AI-generated overview

One Brain is a remote shared-memory server for Claude Code that lets an assistant save and search team decisions, with names and dates attached.

What you can do
Save entries to the team's shared brain, either explicitly or when a topic is closed, with the author's name and date attached to each one. Search the brain for what the team decided about a service, client or person, since every entity is a tag. Hand off session state to a future session or a teammate, and resume where the last session left off. Run a diagnostic command when something is not working. The server stores and retrieves only; no model runs on it.
When to use it
Useful for teams working in Claude Code who keep losing context between sessions and want decisions, rationale and client status shared across teammates without writing documents or maintaining a wiki.
Requirements
A remote endpoint reached over OAuth sign-in with Google or an invited email and password. Claude Code must already be installed, plus a POSIX environment (macOS, Linux, or Git Bash/WSL on Windows); jq is optional. Setup runs a one-line command that installs a plugin, saves an access key and writes rules to the Claude Code configuration file.
Before you install
The setup command saves an access key on the machine and modifies the Claude Code configuration file. The command expires after 15 minutes. Team entries are visible to every member of the brain, so anything recorded is shared; the README states source code is not read and brains are isolated, but review the security and privacy pages before storing sensitive material.

Installation

In SourceWeft

  1. Open One Brain in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.Prophet-git%2Fone-brain/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "one-brain": {
      "type": "http",
      "url": "https://onebrain.prophet.lat/api/mcp"
    }
  }
}

README

One Brain: the memory Claude Code is missing

Every session starts from zero and forgets it all on close. One Brain gives Claude Code permanent memory, shared by the whole team. What one person records, everyone has, with the name and the date still attached.

Create your brain · Free during the open beta · Leer en español

What it looks like

Someone asks their Claude about a service they didn't build:

why does the payments worker stop retrying after 2 attempts?

That cap is deliberate, not a leftover. It was set after the duplicate-charge incident in March.

"Two retries on payments, hard stop. At five we double-charged 40 customers in one night. If a queue genuinely needs more, it comes through me." Priya Raman · Platform · Mar 12

The answer came out of a teammate's session six months earlier. Nobody wrote a document for it.

Why it's different

Nobody writes documents. Memory gets recorded when you close a topic, so there is no form to fill in and no wiki to keep alive.

Every answer has a name and a date on it, which means you stop reconstructing intent from commit messages a year later.

Every service, client or person is a tag. Ask about one and you get everything the team wrote about it, whichever teammate wrote it.

The server itself doesn't run any model. It stores and retrieves, and the judgment stays in your Claude, on your machine.

Getting in

You need Claude Code installed first (how to install it). The command below checks for it, and if it's missing it stops and you have to ask for a new one.

  1. Go to onebrain.prophet.lat and sign in with Google, or with your email and a password if a teammate invited you. Your brain exists immediately, with no form to fill in and nothing to approve.
  2. Pick "Uso Claude Code" (I use Claude Code). The panel shows a one-line command. Paste it in your terminal (on Windows, in Git Bash). It installs the plugin, saves your access key and writes the One Brain rules to ~/.claude/CLAUDE.md. The command works once and expires after 15 minutes; if it does, ask the panel for a new one.
  3. Open Claude Code in any folder and ask it "what does my company's brain know?". The panel shows when the first query arrives.

There are no tokens to copy and no extra restarts. Just open Claude Code after the command has finished: a session that was already open won't see it.

To connect another computer, open Ajustes → Mis computadoras in the panel and run the command it gives you on that machine.

Daily use

Most of the time you do nothing. The team's context arrives when the session starts, and your Claude records what you close.

When you want to be explicit:

  • brain_save, or just tell Claude "save this in One Brain"
  • brain_search for "what did we decide about X?" or "where is client Y at?"
  • /one-brain:handoff to hand off state to your future self or a teammate
  • /one-brain:resume to pick up where the last session left off
  • /one-brain:doctor when something isn't working

Requirements

You need Claude Code already running, either the terminal or the desktop app. One Brain doesn't replace it. It gives it memory.

It also needs a POSIX environment. macOS and Linux already are one; on Windows, use Git Bash or WSL. jq is optional, since the plugin falls back to python3 and then perl.

Working in Codex instead? There's a Codex build of the same brain.

Your data

One Brain stores what your team decides and learns, not your repository. It doesn't read your source code. Every brain is isolated from every other one, you can see and edit every entry, and you can export the whole thing whenever you want without asking anyone.

More on the security and privacy pages.

Support

[email protected] · Built by Prophet

License

Source-available, not open source: see LICENSE. You can read and audit every line before installing it, and run it against the One Brain service. You can't redistribute it or point it at a different service.

Source: README.md at commit d380192

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

1
  1. v0.1.666LatestSep 16, 2026

Score

75% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.

Other Brain integrations

BrainDog

dog.brainOfficial publisher

BrainDog is a hosted record of projects, decisions and money that connected AI agents read before starting work and write their results into.

Streamable HTTPWeb executableAuth
Productivity & WorkflowCredential requiredNov 5, 2026

Codebase Brain

Published by GitHub account Nomannazir

Indexes a local or public git codebase so an assistant can search it, trace git history, check change impact and find stale docs, all with path:line citations.

STDIODesktop only
Web Search & ScrapingNot validated yetNov 7, 2026

Personal Brain

Published by GitHub account zhang-liz

Gives an AI assistant memory of your professional network, stored as a graph in FalkorDB, so it can look up contacts, find people by topic, and save who you…

STDIODesktop only
Databases3 toolsNov 6, 2026

Second Brain

Published by GitHub account ara-5

Shared Graph-RAG memory for AI agents: markdown vault, wikilink recall, tasks, messaging.

STDIODesktop only
Knowledge & Memory12 toolsNov 7, 2026

Make0 Brain

Published by GitHub account make0-official

A local, offline memory server that lets Claude Code, Cursor, Codex and other MCP agents share one versioned store of durable facts.

STDIODesktop only
Knowledge & MemoryNot validated yetNov 7, 2026