Runcell

About Runcell

Runcell is an AI agent that works inside Jupyter rather than beside it. Most notebook AI assistants suggest the next function; Runcell runs the whole loop. Ask it a question and it inspects the notebook, the data, and the existing code, plans the steps, writes the Python, executes the cells, reads the resulting tables and charts, recovers from errors, and carries the outcome forward into the next useful experiment. That closed loop — inspect, plan, execute, read outputs, continue — is what separates it from autocomplete in a notebook, because the next decision is based on what the code actually produced rather than on what the model predicted it would produce. It has three modes of use. The autonomous agent turns a described result into an executed multi-step workflow. In-context assist answers questions about a specific cell, transformation, result, or error, reading the surrounding cells and outputs before applying a fix. Learn-by-doing runs analytical approaches side by side with real outputs so you can compare methods on your own data before committing to one. Alongside the agent it adds conveniences Jupyter has always lacked — a file tree, global search, and git integration — directly in the notebook interface. It installs with pip install runcell. The company behind it is Kanaries Data Inc., which also builds data-exploration tooling.

Key Features

Runcell Pros & Cons

Who Is Runcell Best For?

Tags