Learn

Do it once, start to finish.

Each lesson takes one run from the first request to its result, with the exact requests, commands and fields along the way. Pick the path that matches how you start work.

Build with the API

about 32 min

Start agents from your own server, follow them and act on what comes back.

  1. 01Your first run from codeCreate an org key, start a session with one request, follow it live with curl, read its answer and send a follow-up into the same sandbox.8 min
  2. 02Act on the result with a webhookPoint your key at an endpoint, verify each signed POST, and handle every finished turn exactly once, even when a delivery is retried.8 min
  3. 03Give a run files, skills and connectionsStart a run with files in its sandbox, skills loaded from a URL, the tools it may call and the repository or issue it works from.10 min
  4. 04Run a saved agent from codeSave a harness with its model, instructions, skills and connections once, then start each run with its id and a task.6 min

Run it with your team

about 20 min

Start, watch and steer runs from the console and the terminal, together.

  1. 01Work that starts itselfTurn a prompt into a Project automation that runs on a GitHub event, a Linear issue, a schedule or a webhook, and read what each run did.8 min
  2. 02Steer a run as a teamInvite teammates to your org, see every run in a Project with who started it, queue follow-ups mid-turn and take control of a run's browser.6 min
  3. 03Hand off from the terminalStart a long run from your terminal, close the laptop, and pick it back up later to read what happened, steer it, park it and wake it.6 min