YepCode

Developer Tools

YepCode turns a block of Node.js or Python into something an agent can actually call: you define a reusable process, describe its inputs as a JSON Schema, and YepCode exposes it as a callable tool that runs in an isolated sandbox with npm and PyPI packages available. Neotask connects to that layer directly, so instead of you writing glue code between a script and a chat interface, you ask for the outcome and the agent invokes the right YepCode process with the right arguments. The sandbox isolation means it can run untrusted or heavier logic — a data transform, a webhook responder, a scheduled report — without touching your own machine.

What you can automate

YepCodeDefine, schema, and run reusable Node.js/Python processes as callable AI tools in a sandboxed environment.

Real workflows

Run a custom process from a conversation

You already have a YepCode process that reshapes a CSV export into a billing summary. Ask Neotask to run it against this week's file, and the agent maps your request to that process's schema, executes it in YepCode's sandbox, and returns the output — no manual trigger, no separate dashboard visit.

Frequently asked questions

Does Neotask write the YepCode process for me?

No — YepCode processes are logic you (or your team) define with a JSON Schema for inputs. Neotask's job is calling the process correctly once it exists, matching your request to the right inputs.

What runtime does the code actually execute in?

An isolated sandbox managed by YepCode, with access to npm and PyPI packages, so the process behaves the same whether it runs on demand or on a schedule.