- Overview
- Getting Started with UiPath Agents
- Getting Started with UiPath Agents using LangGraph
- Building a Low-Code Agent in Studio Web
- Adding Tools to Your UiPath Agent
Build and deploy a low-code UiPath agent from the CLI, from scaffolding through Studio Web upload.
In this lab, you will use the UiPath CLI to build a low-code agent from scratch, without opening UiPath Studio until the final step. You will do the following:
- Install the UiPath CLI and the agent tool.
- Install UiPath skills for your coding agent (optional).
- Authenticate to UiPath.
- Scaffold a new solution and low-code agent project.
- Configure the agent by editing
agent.jsonandentry-points.jsondirectly. - Validate the agent and upload it to UiPath Studio Web.
- Test the agent in Studio Web with a real scenario.
There are a few approaches to create UiPath agents. This uses the UiPath CLI; you can also create agents using the low-code Agent Builder and using the LangGraph SDK.
Estimated time: 30-45 minutes
What you are building
This lab walks you through building a Monster Selector, a low-code agent for a fictional Adventurer's Guild in a fantasy role-playing game (RPG) world. The agent picks the most thematically appropriate monster for a quest from a list of candidates. It is deliberately simple so the focus stays on the CLI workflow, not the domain logic. The same agent appears in the companion Adding Tools to Your UiPath Agent lab.
Here is the full design of what you are building:
| Component | Details |
|---|---|
| System prompt | You are an RPG game master. Given a quest description and a list of candidate monsters, pick the ONE whose lore, environment, or threat level best fits the quest. Return only the index slug of your choice. |
Input: questDescription | string; the quest description provided by the caller |
Input: monsters | array; candidate monsters (each with a name and index slug), provided by the caller |
Output: monsterIndex | string; the index slug of the selected monster |
Feel free to adapt the lab.
You can follow along with the Monster Selector example or adapt it to your own use case. To customize it, you only need to adjust three things: the system prompt, your input fields, and your output fields. The rest of the lab works the same way for almost any single-turn classification or selection scenario.
Prerequisites
- CLI version - validated against UiPath CLI v1.1.0. Check your version using
uip --version. Different versions may behave differently; report drift withuip feedback send. - UiPath account - sign up or log in to UiPath Automation Cloud before starting.
- Node.js 18+ - required to install the CLI. Check with
node --version. Download from nodejs.org if needed. - A terminal - PowerShell or Bash both work. The commands in this lab are shell-agnostic.
- Admin rights - installing global npm packages requires elevated permissions. Confirm you can run
npm install -gbefore starting.
No prior UiPath experience is required.
What this lab does not need. Unlike the coded agent lab, this lab does not require Python, Python SDKs, or tooling like uv. And while not required, a coding agent can be a useful building partner. UiPath agents are configured entirely with a system prompt and JSON schemas; no code is required.