- Get started
- Best practices
- Capabilities
- Skills
- CLI
- Examples
- Examples overview
- Recipe: build an agent
- Recipe: automate with RPA
- Recipe: enter data into a web app
- Recipe: extract data from a desktop app
- Recipe: refactor and test a workflow
- Recipe: a queue-based process with REFramework
- Recipe: build a Maestro Flow
- Recipe: coded app and API workflow
- Recipe: verify a release with Test Manager
- Recipe: extract data from documents with IXP
- Advanced
- Help
UiPath for Coding Agents lets you build UiPath automations and agents from inside an AI coding agent such as Claude Code, Cursor, Codex CLI, or Google Antigravity. Instead of switching to a separate designer, you describe what you want in natural language and your coding agent produces, runs, and tests real UiPath projects and solutions.
If you already build with UiPath, this is a new front end for work you recognize. The automations, agents, and platform are the same; the way you create them changes.
The three pieces
Three components work together to make this possible.
- Your coding agent is the assistant you prompt — Claude Code, Cursor, Codex CLI, or Google Antigravity. It reads your instructions and edits files, runs commands, and reasons about your project.
- The
uipCLI (Command Line Interface) is the command-line tool that connects your machine to UiPath. It authenticates to your organization, scaffolds projects and solutions, runs and publishes automations, and exposes the platform to your agent. - UiPath skills are packaged instructions that teach your coding agent how to do specific UiPath tasks, such as authoring an RPA workflow or building an agent. Each skill is maintained in the public UiPath skills repository.
How they fit together
When you ask your coding agent to do UiPath work, it draws on the relevant skill for the rules and patterns of that task, then uses the uip CLI to act against your UiPath organization. The skill supplies the know-how; the CLI supplies the connection and the commands.
The result is that your editor becomes a place where you can go from a plain-language request to a working, runnable UiPath project or solution — with the agent handling the structure and commands, and you steering the intent.
More than building
A coding agent helps across the whole automation lifecycle, not only the first build. The same agent, skills, and uip CLI apply at every stage:
- Discovery and planning — exploring an application, mining a process document, and turning it into a design and a build plan for a single project or a multi-project solution.
- Building — authoring RPA workflows, agents, apps, and API workflows, assembling them into solutions, and working within existing and legacy projects.
- Verification — generating and running evaluations and tests, and running the Workflow Analyzer to catch issues early.
- Troubleshooting — investigating failures from logs and run history, explaining what went wrong, and reading evaluation results to determine improvements.
- Running and managing — packaging and deploying solutions, publishing to Orchestrator, and working with queues, assets, jobs, and platform settings.
This matters because much of an automation's effort and cost lands after the first version ships, in running, maintaining, and operating it. Help in those later phases can matter as much as help at build time. The Skills catalog groups the available skills along these stages.
Who this is for
This guide is written for UiPath builders who are new to AI coding agents. It assumes you understand UiPath concepts and focuses on the coding-agent workflow: installing the tooling, understanding skills, and working across the automation lifecycle.
To set up the tooling, see Install and set up. To see what your agent can do, browse the Skills catalog.
Availability
UiPath for Coding Agents is rolling out capability by capability. Capabilities in Public Preview may change before general availability.
Public Preview:
- Building RPA workflows
- Operating automations
- Building agents
- Building Maestro Flows
- Building coded apps
- Building API workflows
- Building human-in-the-loop tasks
- Testing with Test Manager
- Document Understanding (IXP)
- Troubleshooting
- Platform operations (authentication, Orchestrator, and Data Fabric)
- Managing solutions
- Admin and governance operations
- Audit logs
- Integration Service connections
Disclaimers and responsible use
A coding agent does not validate the compliance of generated code — you are responsible for reviewing it before deploying to production. Read Governance and trust for the full disclaimer and the recommended guidelines before you start.