Quickstart

Quickstart

You need Docker. Nothing else.

git clone https://github.com/usewrit/writ.git && cd writ
./scripts/gen-env.sh
docker compose up -d --build

Open http://localhost:8000 and create your account. That is the whole install.

The first build takes a few minutes. The document/OCR extractor bakes in the OCR runtime and its model weights so extraction works offline and air-gapped.

Then connect one agent

Nothing will execute until an agent is connected. The coordinator holds your data and serves the UI; it never launches a browser itself. Agents do the browsing, and they can run anywhere.

In the app open Fleet → Connect a new agent, copy the line, and run it on whichever machine should do the browsing — your laptop is fine:

curl -fsSL http://localhost:8000/agent.sh | sh -s -- WRIT-4K2P-9XQ

That installs the agent, enrols it and starts it. The pairing code is single-use and expires in 15 minutes. Everything else — the coordinator URL, the document-extractor address and its secret — the installer fetches for itself.

Full detail, including remote agents and manual installs: Connecting agents.

Day-to-day

Run these from the repository root. The root compose.yaml wires in docker/docker-compose.yml and loads your .env automatically.

docker compose logs -f    # follow logs
docker compose restart    # restart
docker compose down       # stop, keep your data
docker compose down -v    # stop and delete everything

What runs

Container	Port	What it does
coordinator	127.0.0.1:8000	API + web UI + scheduler. One SQLite file.
doc-extract	127.0.0.1:8092	Reads PDFs, office files and scanned pages.

Both publish on loopback only. To reach the app from another machine, put it behind TLS — see Production deployment.

First things to try

1. Record a workflow. Workflows → New → Record the steps, give it a URL, and click through the task. Every action becomes an editable step.
2. Replay it. Press Run. It executes on your agent at zero AI-token cost.
3. Put it on a schedule, or expose it as a REST endpoint or an MCP tool.

Next

• Configuration — every setting
• Production deployment — before exposing it to a network
• Troubleshooting — if something is not working