The Lab Agent is a conversational AI built into the Blueprint Editor. Instead of clicking through
the canvas, you describe what you want in plain language and the agent builds or edits the logic for
you — then runs your tests to check its own work. It is the fastest way to make a change when you can
say what you want in a sentence.
The agent works on the whole project: it can create, rename, and delete Blueprints, edit any
section, and read and write your type definitions.
Opening the agent
Open the Agent tab in the right sidebar of the editor.
Type your request in the composer at the bottom — “Add 15 points when the applicant has more than 5 years of employment” or “Explain what this Blueprint does” — and press Enter. The agent streams its response,
showing its reasoning (“Thought for N seconds”) followed by the actions it takes.
The conversation resets when you refresh the page — it is scoped to your current editing session, not
stored as a permanent chat history.
@-mention pills
To point the agent at something specific, @-mention it. Type @ in the composer to reference a
type definition, parameter, variable, or Blueprint, and it is inserted as an inline pill. The agent
also uses these pills in its replies, so there is no ambiguity about which element it means.
Choosing how it works
Two controls in the composer shape how the agent behaves:
| Control | Options | What it does |
|---|
| Mode | Approval / auto-apply | In Approval mode the agent proposes changes and waits for you to review and apply them; the other mode lets it write changes as it goes. |
| Model | Quality / Speed | Quality uses a stronger model for harder logic; Speed trades some depth for faster responses. |
Reviewing and applying changes
In Approval mode, the agent’s edits arrive as a proposed change you can inspect before they touch
your Blueprint. Apply the change to write it in, or discard it to leave your logic untouched. Because the
agent can also run your Blueprints and test suites, it often verifies a change itself and reports the
result inline before handing it back to you.
Attachments
You can attach a PDF or image to a message — hand the agent a requirements document, a spec, or a
screenshot of a rules table and ask it to turn it into logic.
| Tool | Reach for it when |
|---|
| Lab Agent | You can describe the change in a sentence, or want AI to build across multiple Blueprints, run tests, and iterate. |
| AI Editing (Ctrl+I) | You want a quick, scoped edit to the specific node, section, or Blueprint you have selected on the canvas. |
| Inline Editing | You want to type the exact value, title, or expression yourself. |
Tips
- Be specific. “Set the Approve threshold to 65 points” beats “make it stricter.”
- Stay in Approval mode while you are learning what the agent does — you see every change before it lands.
- Let it test. Ask the agent to run your test suite after a change so it catches its own mistakes.
- @-mention to disambiguate when your project has several similarly named elements.
What to do next
