Super Ai Agent
What is Super Ai Agent
The project is a showcase for advanced techniques in developing intelligent agents using Spring AI, SpringBoot 3, and JDK21. It encompasses the implementation of a Manus intelligent agent capable of autonomous tool invocation and decision-making through interaction with a local MCP service.
Use cases
Potential use cases include automated documentation generation, context-aware queries, retrieval-augmented generation (RAG) tasks, and the integration of various data sources for enriched responses. The system can facilitate complex workflows combining local and online resources efficiently.
How to use
To utilize the system, set up a Spring Boot environment with the specified dependencies and configurations. Instantiate the Manus intelligent agent and connect it to the MCP service for tool execution. Use predefined templates and memory mechanisms for customized responses, and incorporate custom query handling and RAG capabilities.
Key features
Key features include custom interceptors for filtering inappropriate content, a specialized memory system for persistent dialogue, versatile document retrieval mechanisms, function call development, and support for both online and offline RAG implementations. The project also supports vector database integrations.
Where to use
This system can be applied in various domains where intelligent automation is needed, such as customer support, information retrieval systems, document automation, and sophisticated conversational agents. Industries like tech, education, and research can benefit from such advanced capabilities.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Overview
What is Super Ai Agent
The project is a showcase for advanced techniques in developing intelligent agents using Spring AI, SpringBoot 3, and JDK21. It encompasses the implementation of a Manus intelligent agent capable of autonomous tool invocation and decision-making through interaction with a local MCP service.
Use cases
Potential use cases include automated documentation generation, context-aware queries, retrieval-augmented generation (RAG) tasks, and the integration of various data sources for enriched responses. The system can facilitate complex workflows combining local and online resources efficiently.
How to use
To utilize the system, set up a Spring Boot environment with the specified dependencies and configurations. Instantiate the Manus intelligent agent and connect it to the MCP service for tool execution. Use predefined templates and memory mechanisms for customized responses, and incorporate custom query handling and RAG capabilities.
Key features
Key features include custom interceptors for filtering inappropriate content, a specialized memory system for persistent dialogue, versatile document retrieval mechanisms, function call development, and support for both online and offline RAG implementations. The project also supports vector database integrations.
Where to use
This system can be applied in various domains where intelligent automation is needed, such as customer support, information retrieval systems, document automation, and sophisticated conversational agents. Industries like tech, education, and research can benefit from such advanced capabilities.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Content
介绍
炫技项目，手搓Manus智能体、MCP server、MCP Client、离线RAG、Function Call等，用代码揭开LLM开发的真面目。
环境
- Spring AI
- SpringBoot 3
- JDK21
智能体实现效果
可以自主调用本地实现的MCP服务中的工具，自主思考，最终给出结果
执行过程，智能体可以自主思考调用什么工具，并根据工具调用返回的结果来决定下一步的策略：
生成PDF结果（markdown格式，图片加载有问题）：
图片链接可以打开：
目前已完成功能
- 自定义拦截器（Advisor）
- 违禁词校验Advisor
- 专用模板类，从文件加载模板
- 自定义ChatMemory
- 结构化输出转换器
- 对话记忆持久化
- 本地RAG
- 在线RAG
- 向量数据库
- 文档检索器
- 上下文查询增强器
- function call开发
- MCP开发（sse和stdio）
- 自定义Manus智能体
TODO
- 权限校验
- 自定义对话记忆，持久化对话到MySQL、Redis中
- 自定义DocumentReader，读取Github仓库信息
- 自定义QueryTransformer查询转换器，使用第三方翻译API代替大模型翻译
- 实现基于向量数据库和其他数据库的混合检索（MySQL、Redis、ES）
- 自动发送邮件tool
- 优化PDF生成，本地存储图片
- 手动控制ToolCallingManager执行流程
- 实现env环境变量传递参数，多agent工作流
Dev Tools Supporting MCP
The following are the main code editors that support the Model Context Protocol. Click the link to visit the official website for more information.
