# AI use cases for Open Intelligent UI (CopilotKit/OpenIntelligentUI)

Repo: [CopilotKit/OpenIntelligentUI](https://github.com/CopilotKit/OpenIntelligentUI) | Use cases: 3 | Page: https://dedupedai.com/use-cases/copilotkit-openintelligentui/

Open Intelligent UI is an open-source generative UI chat app built with CopilotKit and AG-UI that turns questions into interactive explanations, comparisons and working tools, with a Next.js frontend, a Python agent and an optional MCP server.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Run a chat app that answers with interactive UI: https://dedupedai.com/use-cases/copilotkit-openintelligentui/run-generative-ui-chat-locally/
- Adapt the agent and UI to your own workflow: https://dedupedai.com/use-cases/copilotkit-openintelligentui/adapt-generative-ui-to-your-workflow/
- Expose the app's tools through an MCP server: https://dedupedai.com/use-cases/copilotkit-openintelligentui/expose-generative-ui-over-mcp/

## How to run a chat app that answers with interactive UI with Open Intelligent UI

Repo: [CopilotKit/OpenIntelligentUI](https://github.com/CopilotKit/OpenIntelligentUI) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Clone, set up and start the app so questions come back as explanations, comparisons and tools.

**Who it is for:** A product engineer evaluating generative UI for an internal assistant.

**The problem:** Plain text answers are hard to explore and compare.

**The outcome:** A local chat interface that renders interactive answers.

### Steps

1. Run `git clone https://github.com/CopilotKit/OpenIntelligentUI.git` and `cd OpenIntelligentUI`
2. Run `make setup`
3. Set OPENAI_API_KEY in apps/agent/.env or enter keys in the chat header
4. Start everything with `make dev`

### Prompt to paste into your agent

```text
Set up CopilotKit/OpenIntelligentUI with `make setup` and `make dev`, ask it to compare <two options>, and confirm the answer renders as an interactive comparison rather than text.
```

### Caveat

It expects a model plus a visualization router to be connected before answers render fully.

Risk, credentials: Your model API key sits in apps/agent/.env or the chat header, so keep it out of commits.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/copilotkit-openintelligentui/run-generative-ui-chat-locally/

## How to adapt the agent and UI to your own workflow with Open Intelligent UI

Repo: [CopilotKit/OpenIntelligentUI](https://github.com/CopilotKit/OpenIntelligentUI) | Difficulty: Advanced | Time: 2 hours | Basis: From the README

Use the split frontend, agent and design-system packages to customize answers for your domain.

**Who it is for:** A startup engineer building a domain assistant for customers.

**The problem:** Building a generative UI stack from scratch takes weeks.

**The outcome:** A customized fork running your own agent logic and theme.

### Steps

1. Run the frontend with `make dev-app` and the agent with `make dev-agent`
2. Edit the Python Deep Agent in apps/agent for your workflow
3. Adjust the shared theme in packages/design-system
4. Run `pnpm test` and `make lint` before shipping

### Prompt to paste into your agent

```text
Fork OpenIntelligentUI, change the agent in apps/agent to answer questions about <domain>, and run `pnpm test` plus `make lint`. Report any failures.
```

### Caveat

The MCP server in apps/mcp is optional and started separately with `make dev-mcp`.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/copilotkit-openintelligentui/adapt-generative-ui-to-your-workflow/

## How to expose the app's tools through an MCP server with Open Intelligent UI

Repo: [CopilotKit/OpenIntelligentUI](https://github.com/CopilotKit/OpenIntelligentUI) | Difficulty: Advanced | Time: 45 minutes | Basis: From the README

Start the optional standalone MCP server so other agents can use the app's capabilities.

**Who it is for:** An AI engineer who wants the same tools available in an MCP client.

**The problem:** Capabilities locked inside one chat UI cannot be reused by other agents.

**The outcome:** A running MCP server alongside the app.

### Steps

1. Complete setup
2. Start it with `make dev-mcp`
3. Connect your MCP client
4. Call a tool and compare with the chat app result

### Prompt to paste into your agent

```text
Start the OpenIntelligentUI MCP server with `make dev-mcp`, connect <MCP client>, and call one tool. Confirm the result matches what the chat UI shows for the same request.
```

### Caveat

The build and test commands span both pnpm and uv, so you need both toolchains.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/copilotkit-openintelligentui/expose-generative-ui-over-mcp/