Getting started

This guide takes you from a brand-new account to a working project with people and an AI Agent on it. No coding required — everything here happens in the app.

If any of the words below are unfamiliar (workspace, project, task, agent), the Concepts page explains them all in plain language. It’s a useful five-minute read before you start.

New to Retask?

The Free plan has no time limit and needs no credit card. You get room for a small team, a few projects, and a handful of AI Agents to try things out.

1. Create your workspace

A workspace is your whole account — it holds your people, projects, and settings.

1. Go to app.retask.work and sign in.

2. Create a workspace and give it a name (your company or team name works well).

3. You’ll land in the workspace, with the name shown in the switcher at the top.

Note

You can always change the workspace name, color, and description later under Settings → General. See the Workspaces guide.

2. Create your first project

A project groups related work — usually one per app, product, or area.

In the app

Open Projects in the sidebar, click New project, give it a name (for example, “Website”), and you’ll land on an empty Kanban board ready for tasks.

With the CLI

# Authenticate once, then create a project
retask auth login
retask project create --name "Website"

3. Invite your team

Bring in the people you work with. Each person gets a role that controls what they can do — Owner, Admin, Editor, or Viewer.

1. Go to Settings → Members.

2. Click Invite member and enter their email address.

3. Choose a role. Editor is the right choice for most teammates — they can create and change tasks. Use Viewer for people who only need to follow along.

Note

Not sure which role to pick? The Workspaces guide has a table comparing what each role can do.

4. Add an AI Agent

An Agent is an AI worker you can assign tasks to, just like a teammate. You configure it once by filling in a short form — there’s no separate template to wire up.

1. Go to Settings → Members → Agents and click New agent.

2. Fill in the basics: a name, an optional description, and a role — Task processor (acts on a task when its status changes) or Task planner (plans and assigns tasks automatically).

3. Choose when it runs and where: a trigger (for example, “when a task is assigned to this agent”) and an execution target (such as “run on any available sandbox”). Add any integrations it needs, and an optional agent prompt.

4. Save, and the agent appears alongside your human members, ready to be assigned work.

Note

The Advanced tab lets you fine-tune the environment — a sandbox startup command, a shutdown policy, and the session’s system prompt — but the defaults are a fine place to start. See the Agents guide for the full set of options.

Caution

Agents do real work on your code inside a sandbox. Start them on a low-stakes task first so you can watch what they do before trusting them with anything important.

5. Create a task and assign it

Now put it all together: add a task and hand it to a person or your new agent.

In the app

Click + Task at the top, fill in a title and description, set a priority, and choose an assignee — a teammate or your AI Agent. Save, and it appears on the board.

With the CLI

retask task create \
  --project-id <id> \
  --title "Fix login bug" \
  --priority HIGH

To have an AI Agent do the work, open the task and use the Task assistant menu — Plan this task or Process this task. The agent runs inside an isolated sandbox and opens a session you can watch live. The full flow is covered in Run an AI Agent on a task.

Where to go next

• Read the Concepts page for the mental model behind everything.
• Walk through the headline flow: Run an AI Agent on a task.
• Dig into Workspaces, Projects, and Tasks.
• Automate everything from the terminal with the Retask CLI.