Learn to work with Refactor
Refactor is where your team does repeatable client work with AI agents. The agent works in a workspace beside your files, a person checks what matters, and what you learn carries into the next piece of work.
New here? Read these three first
- What Refactor is
The idea in five minutes: workspaces, the agent, and a person checking what matters.
- Quickstart
Create a workspace, give the agent a piece of work and review the result.
- Find your way around
Home, the sidebar and Needs you: where to look when you sign in.
Browse by area
Get started
What Refactor is, your first working result, and the words the console uses.
Workspaces
Where the work happens: files, a conversation with the agent, and the tools around them.
Fibo
Your assistant across all your work: what needs you, what it remembers, how it reaches you.
Skills
Written instructions the agent follows, kept in one place, versioned, tested and reviewed.
Services and engagements
Repeatable work as a service, and each piece of work done for someone as an engagement.
Data and connections
The data, client folders, repositories and sites your work reads and writes.
Administration
People and access, which models are allowed, your own settings, and how data is protected.