Skip to content

Repository files navigation

Company Brain

Company Brain

A teammate in your Slack that truly knows and understands your company, and can do anything.

Deploy · User guide · Permissions · Use cases · Discord

Apache 2.0 Memory by supermemory Runs on Cloudflare Workers

Used to be a paid product with thousands of users. Now it's free and open source.
Read the announcement →

Deploy to Cloudflare


A few weeks ago, we discontinued our Company Brain product at supermemory. This is that product, the whole thing, open sourced and rebuilt to run on your own Cloudflare account.

Your team's knowledge is scattered across Slack threads, docs, tickets and people's heads. Company Brain sits in Slack, remembers what your team says, and answers from it. Then it goes and does the work: opens the issue, reads the PR, digs into the repo. And when a conversation needs something it knows, it speaks up on its own.

#eng

Alex: is prod down? a couple of customers are pinging me

Company Brain: Not fully down. api/search is elevated, 42 errors in the last 15 minutes, and Kush is already on it. Probably this morning's deploy.

Nobody mentioned the bot. It spoke up because it knew the answer.

🧠 Remembers Decisions, projects, owners and context from the channels it's in, kept current as people talk. No one has to write anything down.
💬 Answers from your team Not from the internet, not from guesswork: from your own conversations and connected tools.
🛠️ Acts in your tools GitHub, Linear, Notion, Google Workspace and hundreds more over MCP. It opens the issue instead of telling you to.
📣 Speaks up Chimes in when a conversation needs something it knows. You decide how chatty it is, org-wide and per channel.
💻 Runs code Its own sandbox: clones repos, runs scripts, and hands back charts, CSVs and PDFs right in Slack.
🌙 Works while you sleep Scheduled digests to a channel or your DMs, plus research it kicks off on its own.
🎓 Learns your way Skills teach it your processes, formats and voice. A workspace prompt sets how it behaves everywhere.

Private by design

Memory isn't one big bucket. It's a permissions graph, and the brain only ever reads with the asker's own access.

Where you ask What it can draw on
A public channel The shared brain everyone in the org can see
A private channel That channel's memory, plus the shared brain
Your DMs Your personal memory, every private channel you're in, and the shared brain

So it can't leak something you couldn't see yourself. Tool access works the same way: writes always run under your own connection, and when a request needs a tool only a teammate has connected, it asks them first with an approve or deny card. Nothing is granted silently.


Get started

💬 I want it in my Slack

One click to deploy, two API keys, and a setup page that walks you through the Slack app. The database sets itself up.

→ Deploy in five minutes

🔧 I want to hack on it

TypeScript on Cloudflare Workers, Durable Objects and D1, with memory on supermemory. Run it all locally.

→ Local development


Deploy in five minutes

1. Click deploy. It asks for two secrets:

  • SUPERMEMORY_API_KEY: where the brain keeps its memory. Get one at console.supermemory.ai.
  • MODEL_API_KEY: an Anthropic, OpenAI, Google or xAI key, whichever you have. The brain works out the provider from the key. You pay the provider directly, no markup. An OpenRouter key works too, and reaches every provider's models through one account.

Everything else is provisioned for you: D1, KV, Durable Objects and Workers AI.

2. Open /setup on your new worker. It checks your keys, hands you a Slack app manifest with your URLs already filled in, and takes the Slack credentials back.

3. Sign in with Slack. The first person to sign in owns the deployment.

4. Install to Slack, and say hi. 👋

Your team signs in with Slack at / to see the brain's home, a live graph of everything it remembers, and settings for tools, models, proactivity, automations and skills.

Free plan or Workers Paid?

Company Brain runs on Cloudflare's free plan. Workers Paid ($5/mo) is better, and worth it if your team leans on the brain:

Free Workers Paid
Long, multi-step answers Can get cut short: the free plan allows 50 outbound calls per request Room for the brain's full tool loop
Code sandbox (shell, git, Python) With a Daytona key (DAYTONA_API_KEY) Built in, on a Cloudflare container. Daytona still works if you prefer it.

Everything else works the same on both, including Code Mode: when the brain writes code to chain tool calls or crunch your team directory, it runs in QuickJS inside the worker, not on paid Dynamic Workers.

To use the built-in container on Workers Paid, uncomment the Workers Paid block in wrangler.jsonc, set CONTAINER_SANDBOX to "on", and redeploy.


Local development

bun install
cp .dev.vars.example .dev.vars   # fill in the two keys
bun run dev

If you've enabled the Workers Paid block, Docker has to be running for the sandbox container. Slack has to reach your machine, so point a tunnel at the dev server, set PUBLIC_URL in .dev.vars to the tunnel's URL, and create the Slack app from the tunnel's /setup page.

After changing the schema in src/db/schema, run bun run db:generate. It writes the migration and bundles it into the worker, which applies it on its next request.


User guide · Architecture notes · supermemory

Built by supermemory. Apache 2.0.

About

Open-sourcing our company brain - A teammate in your Slack that remembers everything your team says, and can go do the work.

Resources

Stars

505 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages