Coding a Multi-Agent Game Master with Spring AI
summary
Live-coded a multi-agent Game Master that runs end-to-end: Spring AI + Java 25 + Gbang for agents, Amazon Bedrock (CloudIQ-4.5) as the LLM, and Spring Boot for MCP tool servers. Implements a custom dice tool, exposes it over an MCP server, and connects specialized agents via MCP/A2A protocols using JSON “cards” for discovery. Demo shows MCP inspector, a dice roll over the network (examples: 1d20 = 7; 3d20 = 14, 1, 3), character and rules agents cooperating, and a working orchestrator; repo provided. MCP inspector
AI-written summary · about
chapters (10)
- 0:00AI game master with Spring AI
- 3:24Agentic AI and MCP protocols
- 6:54Building AI game master agent
- 16:05Built-in web fetching tools
- 20:19Custom dice rolling tools
- 24:55Scaling with MCP server
- 33:08Specializing agents with A2A
- 37:04Integrating specialized agents
- 39:13Creating AI game master character
- 41:41Summary and resources
related talks
27mBuilding AI Agents with Spring & MCP
54mFrom Single-Shot LLMs to Intelligent Agents: Building Scalable AI Systems with Spring AI and MCP
45mAgents, Tools, and MCP, oh my! Next-level AI concepts for developers
52mBeyond local tools: Deep dive into MCP with Spring AI
01h 03mAI Coding Agents and how to code them
32mFrom Monolith to AI Agent Modernizing Java Systems with MCP
49mBuilding AI Agents with Spring & MCP
24mBuilding Agentic Applications with Spring AI