How it works
Someone on your team sets up the brain once (that part uses the terminal). After that, you open Cowork and your company context is just there. Product knowledge, customer info, team processes - loaded into every session automatically.
What you get
Full company context
Every conversation starts with your product, customers, positioning, and team knowledge. No pasting, no re-explaining.
Shared skills
Say "prep me for the Acme call" or "write a cold email to CTOs" and the right workflow runs. Same skills your engineering team uses.
Live connections
Claude can check your calendar, look up contacts in your CRM, pull invoices from Stripe - all from Cowork. No switching apps.
Always current
When someone on the team updates the brain, your next Cowork session picks it up automatically. No manual syncing.
Setup
There are two paths depending on your comfort level:
Have a developer set it up
Ask someone on your team to run the setup. It takes 5 minutes. They'll run:
npx @antidrift/cli initThen connect it to Cowork with:
antidrift connect google --coworkAfter that, open Cowork. Your brain is loaded. You're done.
Join an existing brain
If your team already has a brain set up, you just need to clone it:
npx @antidrift/cli join org/repoThis is a one-time terminal command. After that, Cowork loads the brain automatically every time you open it.
What you can say
Once the brain is loaded, just talk to Claude naturally:
Claude knows your company context, your brand voice, your customers, and your team's processes. No setup per conversation.
Connections that work in Cowork
When your team connects services with the --cowork flag, they're available in your Cowork sessions too:
Ready?
Ask your developer to set it up, or run the join command yourself. One terminal command, then Cowork handles the rest.
npx @antidrift/cli join org/repo