modelcontextprotocol/java-sdk
Watch
1
0
Fork
You've already forked java-sdk
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Refactor the mcp-core module

Browse source 
- The core module has been refactored to remove unnecessary abstraction
  between the SSE transport and the STDIO transport.
- Tests have been improved and no longer complain about multiple
  subscriptions to single-subscriber Sink.
- SSE transport layer is now assuming non-blocking execution and no longer
  makes thread hops when not needed.
- Decouple ObjectMapper and error handling from AbstractMcpTransport
  and remove later.

Additional improvements:

- Rename transport classes to better reflect client-side role (*ServerTransport -> *ClientTransport)
- Introduce fluent builder pattern for client configuration
- Simplify DefaultMcpSession by removing ObjectMapper dependency
- Move unmarshal responsibility to transport layer
- Update documentation and diagrams to reflect new architecture
- Clean up test implementations
This commit is contained in:
Dariusz Jędrzejczyk 2024-12-17 18:10:46 +01:00 • committed by Christian Tzolov
commit 7a98a935af
1 changed files with 560 additions and 761 deletions
Download patch file Download diff file Expand all files Collapse all files

18
README.md
View file

@ -1,10 +1,10 @@
# Spring AI MCP
Java SDK for the Model Context Protocol (MCP), providing seamless integration between Java and Spring applications and MCP-compliant AI models and tools.
Java SDK for the Model Context Protocol (MCP), providing seamless integration between Java and Spring applications and MCP-compliant AI resources and tools.
## Overview
Spring AI MCP is an experimental project that provides Java and Spring Framework integration for the [Model Context Protocol](https://modelcontextprotocol.org/docs/concepts/architecture). It enables Spring applications to interact with AI models and tools through a standardized interface, supporting both synchronous and asynchronous communication patterns.
Spring AI MCP is an experimental project that provides Java and Spring Framework integration for the [Model Context Protocol](https://modelcontextprotocol.org/docs/concepts/architecture). It enables Java applications to interact with AI models and tools through a standardized interface, supporting both synchronous and asynchronous communication patterns.
<img src="spring-ai-mcp-architecture.jpg" width="600">
@ -12,22 +12,20 @@ Spring AI MCP is an experimental project that provides Java and Spring Framework
The project consists of two main modules:
### [spring-ai-mcp-core](./spring-ai-mcp-core/README.md)
### [mcp-core](./spring-ai-mcp-core/README.md)
The core module provides a Java implementation of the Model Context Protocol specification. It includes:
- Synchronous and asynchronous client implementations
- Standard MCP operations support (tool discovery, resource management, prompt handling)
- Stdio-based server transport
- Reactive programming support using Project Reactor
- Standard MCP operations support (tool discovery, resource management, prompt handling). Support for request and notificaiotn handling.
- [Stdio](https://spec.modelcontextprotocol.io/specification/basic/transports/#stdio) and [SSE](https://spec.modelcontextprotocol.io/specification/basic/transports/#http-with-sse) transport implementations.
- [Find more](./spring-ai-mcp-core/README.md).
[find more](./spring-ai-mcp-core/README.md)
### [spring-ai-mcp-spring](./spring-ai-mcp-spring/README.md)
### [mcp-spring](./spring-ai-mcp-spring/README.md)
The Spring integration module provides Spring-specific functionality:
- Integration with Spring AI's function calling system
- Spring-friendly abstractions for MCP clients
- Automatic conversion between JSON and Java objects for tool arguments
- Auto-configurations (WIP)
## Requirements