Spring AI Playground is a self-hosted web UI for low-code AI tool development with live MCP server registration. It includes MCP server inspection, agentic chat, and integrated LLM and RAG workflows, enabling real-time experimentation and evolution of tool-enabled AI systems without redeployment.
# Add to your Claude Code skills
git clone https://github.com/spring-ai-community/spring-ai-playgroundSpring AI Playground is a self-hosted web UI platform for building low-code tools and dynamically exposing them via built-in MCP server for AI agents.
Unlike most AI playgrounds that focus solely on prompt testing and chat visualization, it bridges the gap between static AI conversations and real-world actions by enabling you to create executable tools that AI agents can use.
It brings together Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and low-code tool development in a single environment. Tools created in the Tool Studio are dynamically evaluated and loaded at runtime, then automatically made available as Model Context Protocol (MCP) tools. This makes them instantly available to MCP-compatible clients without restarting or redeploying.
<p align="center"> <b>Agentic Chat Demo</b><br/> Tool-enabled agentic AI built with Spring AI and MCP </p> <p align="center"> <a href="https://youtu.be/FlzV7TN67f0"> <img src="https://img.youtube.com/vi/FlzV7TN67f0/0.jpg" width="800"/> </a> </p>Create AI tools using JavaScript (ECMAScript 2023) directly in the browser. Powered by GraalVM Polyglot, these tools run inside the JVM (Polyglot) with configurable security constraints and are immediately exposed via built-in MCP Server. Experience a no-restart, no-redeploy workflow: just write, test, and publish.
Connect to external MCP servers, inspect available tools, and validate tool execution behavior. Test both your custom Tool Studio tools and third-party MCP services in a unified interface.
No comments yet. Be the first to share your thoughts!
Upload documents, configure chunking and embeddings, and test retrieval pipelines. Evaluate prompt execution against selected RAG sources to validate knowledge-grounded responses.
Interact with LLMs in a chat interface where models can reason, select tools, and execute actions. Combine MCP tools and RAG-enhanced context to validate end-to-end agent workflows.
Spring AI Playground is intentionally built as a tool-first reference environment for exploring, validating, and operationalizing Spring AI features in a reproducible way.
Note: This project is intentionally opinionated and scope-limited in its early stages. It foc...