Working with the agent

The concepts page covers what the agent is. This is about getting good work out of it.

Ask for outcomes, not keystrokes

The agent is better at "what" than "exactly how." Tell it the result you want:

"Find the columns most correlated with churned and show the top ten."

That beats spelling out df.corr(), sort, .head(10). It lets the agent handle the edge cases (non-numeric columns, NaNs) without you listing them. Save the precise instructions for when the details actually matter to you.

Give it the context it needs

The agent acts on what it knows. If a result depends on something specific, like which column is the target, what "recent" means, the units, or which file you mean when there are several, say so up front. When it has to guess, it'll either ask or pick a default, and telling it early saves a round-trip.

Reading the chat

The agent's work comes through as cards:

• Plan: the steps it's about to take. Reply to redirect it.
• Thinking: collapsible reasoning, on the models that support it.
• Suggestion: a proposed cell or edit, with Accept, Edit, and Skip.
• Tool call: a compact row for something it ran. Expand for details.
• Clarification: a question it needs answered.
• Research: web results, with sources.
• Sub-agent: a worker it spawned for part of the job.

Stepping in

You don't have to wait for it to finish. Send a new message mid-run and it folds your input in. Edit a cell yourself and it continues from your version. Skip a suggestion and it drops it, and since it remembers your reactions, skipping the same kind repeatedly trains it to stop offering them.

When something goes wrong

If a cell errors, the agent usually reads the traceback and tries a fix on its own. If it loops or heads the wrong way, stop it and be specific about what's off. "The error is a dtype mismatch on the join key" gets a fix faster than "that didn't work."

Attach files for context

You can attach a file to a chat message, like a data dictionary, a spec, or a sample, so the agent has it while it works. For datasets you'll actually load and run against, upload them to the workspace instead, so the code reads the real file.

Match the model to the task

A quick reformat doesn't need a frontier model, but designing an approach might. You can change the model and reasoning effort per message, so it's worth matching them to the work, both for quality and for cost.