Agents-Flex构建Java AI应用

Agents-flex is A lightweight Java AI agent development framework (positioned as a counterpart to Spring AI). It supports features such as…

English | 简体中文 Agents-Flex Agents-Flex is a lightweight AI application development framework for the Java ecosystem. It organizes LLM calls, Tool Calling, Agents, RAG, vector stores, Embedding, image generation, audio, MCP, Skills, Text2SQL, and related capabilities into clear modules, so developers can compose only what they need without being locked into a specific runtime or application framework. It is suitable for building intelligent customer service, enterprise knowledge bases, natural-language data analysis, Agent workflows, model gateways, AI-assisted office tools, plugin-based tool systems, and Java services that need to connect to multiple model providers at the same time. Highlights Java native: Core modules are compatible with Java 8+ and can run in plain Java, Spring Boot, or other JVM stacks. Unified model abstractions: ChatModel, EmbeddingModel, ImageModel, RerankModel, and other interfaces wrap provider-specific capabilities. Consistent sync and streaming APIs: The same Prompt, Options, interceptor, and context mechanisms work for both normal chat and streaming output. Complete Tool Calling flow: Supports annotation-based scanning, programmatic tool building, tool execution, tool message feedback, and tool-level observability. Durable Agent runtime: Provides snapshot recovery, tool approval, version-safe persistence, middleware, and event streams. Full RAG building blocks: Includes document models, parsing, splitting, Embedding, vector stores, retrieval, and Rerank support. Production-oriented design: Includes model routing, retry, load balancing, circuit breaking, OpenTelemetry observability, and Text2SQL safety interceptors. Modules Module Description agents-flex-core Core abstractions: Chat, Prompt, Message, Tool, Memory, Document, Store, and observability agents-flex-async-task Durable tracking and scheduling for provider jobs that use submit-and-poll workflows agents-flex-async-task-store JDBC and Redis persistence for async task state, scheduling indexes, and Worker leases agents-flex-doc-extractor Extracts Markdown-style content from PDF, Office, HTML, email, archives, streams, and URLs agents-flex-agent Durable Agent runtime: Run state, snapshot recovery, version CAS, approval, middleware, and events agents-flex-agent-store JDBC and Redis persistence for Agent runs, commands, events, and artifacts agents-flex-observability OpenTelemetry span and metric exporters with JDBC persistence agents-flex-chat Chat model integrations: OpenAI-compatible APIs, Qwen, Ollama, DeepSeek, LiteLLM, llmman agents-flex-embedding Embedding model integrations: OpenAI, Ollama, Qwen agents-flex-image Image model integrations: OpenAI, Gemini, Qwen, Alibaba Cloud, Gitee, Qianfan, SiliconFlow, Stability, Tencent, Volcengine agents-flex-video Asynchronous video generation and editing: Alibaba Cloud Model Studio and Volcengine Ark agents-flex-audio Speech-to-text and text-to-speech: Alibaba Cloud, Tencent Cloud, Volcengine agents-flex-ocr OCR model integrations: Baidu AI Cloud, Gitee AI, and MinerU agents-flex-store Vector stores: Redis, Qdrant, Chroma, Pgvector, MariaDB, Milvus, OpenSearch, Elasticsearch, Alibaba Cloud, Tencent Cloud agents-flex-graph Portable graph SDK, Neo4j/Nebula backends, knowledge extraction, and progressive knowledge-query tools agents-flex-search-engine Search engine wrappers: Lucene, Elasticsearch, and search service interfaces agents-flex-rerank Rerank models: default implementation and Gitee Rerank agents-flex-tool Common tools: file system, Shell, Grep, Glob, WebFetch, Python, JavaScript agents-flex-toolsearch Progressive tool discovery with in-memory and custom search providers agents-flex-mcp MCP client that converts external MCP tools into Agents-Flex Tool instances agents-flex-skills File-system-based Skills loading with progressive disclosure agents-flex-skills-artifact Versioned Skill package storage backed by file systems and object storage services agents-flex-skills-sandbox Sandbox runtime integrations for isolated Skills execution agents-flex-skills-open-sandbox Isolated Skills execution through the OpenSandbox runtime agents-flex-skills-aio-sandbox Isolated Skills execution through an AIO Sandbox service agents-flex-subagent Subagent definitions, background task execution, and output retrieval tools agents-flex-text2sql Natural-language data analysis tools with progressive schema disclosure, read-only SQL checks, and interceptor chains agents-flex-websearch Web search tools with Brave, Bocha, Baidu Qianfan, and custom search providers agents-flex-wiki LLM Wiki support: organize knowledge as a navigable hierarchical Wiki tree with path-based recursive reading and progressive disclosure agents-flex-spring-boot-starter Spring Boot auto-configuration for common models and vector stores demos Example projects Requirements Most modules: JDK 8+ agents-flex-mcp: JDK 17+ Build tool: Maven The current repository version is defined by the revision property in the root pom.xml; it is currently 2.3.1. Installation For plain Java projects, you can use the aggregate dependency: <dependency> <groupId>com.agentsflex</groupId> <artifactId>agents-flex-bom</artifactId> <version>2.3.1</version> </dependency> For Spring Boot projects, use the starter: <dependency> <groupId>com.agentsflex</groupId> <artifactId>agents-flex-spring-boot-starter</artifactId> <version>2.3.1</version> </dependency> You can also depend on only the modules you need: <dependency> <groupId>com.agentsflex</groupId> <artifactId>agents-flex-chat-openai</artifactId> <version>2.3.1</version> </dependency> <dependency> <groupId>com.agentsflex</groupId> <artifactId>agents-flex-store-redis</artifactId> <version>2.3.1</version> </dependency> Quick Start The following example uses an OpenAI-compatible API. Replace endpoint, model, and apiKey with your own model service configuration. import com.agentsflex.core.model.chat.ChatModel; import com.agentsflex.model.chat.openai.OpenAIChatConfig; public class ChatDemo