UX Buddy - AI powered pair designer

About this plugin

UX Buddy is your AI-powered design partner, built to streamline your UX workflow and enhance your design process within Figma.

1. Uncover Assumptions: Identify and challenge underlying assumptions in your designs.
2. Discover Edge Cases: Proactively find potential error states and unusual user interactions.
3. Evaluate Usability: Get AI-driven feedback on the usability of your selected layers based on established heuristics.
4. Check Accessibility: Perform quick accessibility checks, considering WCAG guidelines for aspects like color contrast and more.
5. Contextual Analysis: Provide persona details, research insights, and other context to get more tailored feedback from your AI Buddy.
6. Flexible AI Providers: Connect to various AI services including OpenAI, GroqCloud, Hugging Face, or your local Ollama instance.
7. Customizable Prompts: Tailor UX Buddy to your specific needs by creating and managing your own prompt library.

Whether you're brainstorming, refining user flows, or conducting detailed evaluations, UX Buddy acts as an intelligent assistant, helping you make more informed design decisions and create better user experiences, faster.

--------

How to set up AI Provider:

☁️ Option 1: Cloud APIs (OpenAI, GroqCloud, Hugging Face)

Use cloud-based AI models by connecting to one of these supported services.

1. Select “Cloud API”.
2. Choose your service (OpenAI, GroqCloud, or Hugging Face) from the dropdown.
3. Enter your credentials:
4. OpenAI: API Key from platform.openai.com
5. ⚠️ Important: OpenAI requires a funded account to use the API. Free trial credits may be available, but ongoing usage requires a billing setup.
6. GroqCloud: API Key from console.groq.com
7. Hugging Face: Access Token from huggingface.co/settings/tokens
8. Enter the Model Name : choose a vision-capable model:
9. OpenAI: e.g., gpt-4o, gpt-4-vision-preview
10. GroqCloud: e.g., meta-llama/llama-4-scout-17b-16e-instruct
11. Hugging Face: e.g., google/gemma-3-27b-it
12. Click “Test Current AI Connection” to verify everything works.

-----

💻 Option 2: Ollama (Local AI)

Use Ollama to run models locally on your machine—no internet required once models are downloaded.

🔧 Prerequisites:

1. Install Ollama: Download it from ollama.com.
2. Run Ollama:
3. Launch Ollama via terminal (ollama serve) or the desktop app.
4. For convenience, you can set it up to run in the background automatically:
5. macOS: via a plist launch agent
6. Linux: via a systemd service (Note: These are part of your system configuration and not managed by UX Buddy.)
7. Download Your Model:
8. Use ollama pull llava or another vision-enabled model.
9. Ensure the model supports your desired functionality (e.g., image analysis).

🛠️ Configuration in UX Buddy:

1. Select “Ollama” as your provider.
2. Enter your API Base URL — usually http://localhost:11434.
3. Enter the Model Name (e.g., llava, llama3, etc.).
4. Click “Test Current AI Connection”.