agents-flex

Active
GitHub Java Apache-2.0

Description

Agents-flex is A Lightweight Java AI Application Development Framework.

Key Features

  • Lightweight modular Java framework for LLM application development with Apache 2.0 license
  • Native MCP (Model Context Protocol) support for standardized external tool and data source integration
  • AI Skills system for encapsulating business capabilities into reusable, composable Skill units
  • Built-in Text2SQL engine enabling natural language data querying without code
  • Comprehensive model integration supporting mainstream LLMs, Ollama local deployment, and HTTP/SSE/WS protocols
  • Enterprise-grade features including sensitive data masking, OpenTelemetry observability, and distributed tracing

Use Cases

💡 Enterprise-grade LLM chat applications with multi-turn memory and streaming responses
💡 Building AI agent frontends with function calling and MCP tool orchestration
💡 Natural language to SQL data analysis for business users without coding
💡 RAG (Retrieval-Augmented Generation) systems combining embedding, vector store, and re-ranking
💡 Production deployment with Spring Boot integration and configuration center management

Categories

Quick Start

Add Maven dependency `com.agentsflex:agents-flex-core:2.1.6` (and optional modules like `agents-flex-mcp`). Configure a model provider (OpenAI/Ollama/GiteeAI), then call `chatModel.chat()` for sync, streaming, or async inference. For production, use `agents-flex-spring-boot-starter` with encrypted config for API keys.

Related Projects