UiPath for Coding Agents lets you bring your favourite coding agent — Claude Code, Codex, Gemini CLI, anything — and use it to build, run, and ship real UiPath automations. Your coding agent gets a set of UiPath skills and the uip CLI, so it works the way a senior RPA developer would: REFramework, Object Repository, strict selectors (wnd / html / webctrl / aa / uia / java / sap), Try/Catch + RetryScope, Config.xlsx over hardcoded values. So instead of asking your coding agent for answers… you let it build.
Get started
- Install the uip CLI
npm install -g @uipath/cli - Run
uip skills install— or pick the skills you need from GitHub - UiPath/skills: This is a repository of skills for interfacing UiPath capabilities to external developers. · GitHub
The Challenge
Pick your coding agent and use it to take a UiPath automation from idea to running. Not just code suggestions. Not just snippets. An actual project that runs end-to-end.
Example ideas
- Build from a prompt — describe the process in plain language and watch your coding agent scaffold a working project
- Fix and refactor — point it at a failing workflow, let it debug, refactor, harden it (try/catch, retries, arguments instead of hardcoded values)
- Coded agent end-to-end — build a Python LangGraph (or LlamaIndex / OpenAI Agents) agent, run it locally, push it to Orchestrator
- Low-code agent lifecycle — build, run, evaluate, deploy a low-code agent from your terminal
- Deploy + operate — your coding agent handles the full publish → deploy → monitor → roll back loop on Orchestrator
Technical angle
- Use any coding agent that runs in your terminal — including Studio Desktop’s built-in terminal (Claude Code, Codex, Gemini CLI, anything)
- uip CLI + UiPath skills are the integration layer
- Show how your coding agent:
- calls UiPath skills across the lifecycle (build, run, publish, deploy, Orchestrator ops, diagnostics)
- works through the project end-to-end
- recovers when something fails (a faulted job, a publish error, a broken selector)
What to share
- what got built and which coding agent powered it
- if you customised anything — custom skills, system prompt tweaks, agent-specific config — show what and why (no need to customise, but if you did, we want to see it)
- a short demo (video or screenshots) — bonus if it shows the agent recovering from a real failure
- a key snippet or repo
- one thing that surprised you
Bonus points
- Multi-step reasoning — your coding agent chains build → run → debug → fix → deploy → monitor without losing context
- Multi-agent setups — multiple coding agents with clear roles (e.g. Claude Code plans, Codex builds, Gemini CLI tests)
- Human-in-the-loop on risky steps — your coding agent pauses before hard-to-reverse actions (e.g. publishing to a production Orchestrator folder)
- Skills mastery — using a wide range of UiPath skills, or extending your coding agent with custom ones you wrote
Goal
Explore what happens when coding agents don’t just suggest UiPath code… they build and ship real UiPath work.
Drop your builds, experiments, or even partial attempts.
If something breaks, run /uipath-feedback from inside your coding agent.
What is the timeline?
May 5 — Challenge opens
May 15 May 22 — Deadline to submit your functional use case
May 18-21 May 23 - 28— Jury evaluation period
May 22 May 29 — Winners announced 
Please note: while you are welcome to submit multiple demos, only one prize will be awarded per individual. If you contribute more than one solution, the judging panel will evaluate all your entries, and the highest-rated one will be considered for a prize.
What are the prizes?
1st Place — $100 gift card
2nd Place — $80 gift card
3rd Place — $70 gift card
4th Place — $60 UiPath Swag voucher
5th Place — $50 UiPath Swag voucher
What are the judging criteria?
1. Innovation & Idea (40%) How original is the approach? Does it use UiPath skills and coding agents in interesting ways? Does it shift how others would approach their own automation work?
2. Enterprise Relevance & Applicability (30%) Can this realistically be used in a business context? Is the use case practical, scalable, and valuable for real-world automation?
3. Integration Quality & Demo Completeness (30%) Does the demo show a complete, end-to-end flow from input to outcome
How to submit your functional use case?
- Submit your demo to the Forum under the submission type “UiPath for Coding Agents” HERE. Only submissions posted this way will be considered.
- Provide all required information as outlined in the form. Incomplete submissions cannot be evaluated by the jury.
- Verify access to all files and links. Reviewers must be able to view and download your materials without restrictions
A big shout-out to @alexandru and @Simona_Boboc from the UiPath Product Team, thanks for teaming up with us and helping shape this challenge for the whole community, and for being part of the use cases judging! 
.
Good luck, UiPath Community!
Every contribution matters, so let’s inspire each other, learn from one another, and celebrate the amazing talent in our community. 
Don’t forget to spread the word about the UiPath for Coding Agents challenge on social media!! 
Let’s hype this challenge, get more brilliant minds involved, and make it an unforgettable showcase of innovation. 
By submitting the demo, you agree to the challenge Terms & Conditions below 
.
2026.05.5_UiPath_Contest_Regulation_UiPath for Coding Agents.pdf (235.6 KB)
.