diff --git a/README.md b/README.md index 4596816f5..68c62d209 100644 --- a/README.md +++ b/README.md @@ -8,23 +8,25 @@ The EinoExt project hosts various extensions for the [Eino](https://github.com/c - **component implementations**: official implementations for Eino's component types. -| component type | official implementations | -|----------------------|----------------------------------------| -| ChatModel | OpenAI, Claude, Gemini, Ark, Ollama... | -| Tool | Google Search, Duck Duck Go... | -| Retriever | Elastic Search, Volc VikingDB... | -| ChatTemplate | DefaultChatTemplate... | -| Document Loader | WebURL, Amazon S3, File... | -| Document Transformer | HTMLSplitter, ScoreReranker... | -| Indexer | Elastic Search, Volc VikingDB... | -| Embedding | OpenAI, Ark... | -| Lambda | JSONMessageParser... | +### Component Summary + +| Component Type | Count | Official Implementations | +|----------------|-------|--------------------------| +| **ChatModel** | 10 | OpenAI, Claude, Gemini, Ark, ArkBot, Ollama, Qwen, Qianfan, DeepSeek, OpenRouter | +| **Embedding** | 8 | OpenAI, Ark, Gemini, Ollama, Qianfan, DashScope, TencentCloud, Cache | +| **Indexer** | 10 | Elasticsearch (7/8/9), OpenSearch (2/3), Milvus, Milvus2, Qdrant, Redis, Volc VikingDB | +| **Retriever** | 12 | Elasticsearch (7/8/9), OpenSearch (2/3), Milvus, Milvus2, Qdrant, Redis, Dify, Volc VikingDB, Volc Knowledge | +| **Tool** | 10 | Bing Search, DuckDuckGo, Google Search, Wikipedia, SearXNG, BrowserUse, Command Line, HTTP Request, MCP, Sequential Thinking | +| **Prompt** | 2 | CozeLoop, MCP | +| **Document** | 12 | File/URL/S3 Loaders, HTML/PDF/DOCX/XLSX Parsers, Recursive/HTML/Markdown/Semantic Splitters, Score Reranker | +| **Callback Handler** | 4 | APMPlus, CozeLoop, Langfuse, Langsmith | + +📋 **For detailed component information (import paths, descriptions, features, and GitHub links), see [components.md](components.md)** For more details about component types, please refer to the [Eino component documentation.](https://www.cloudwego.io/zh/docs/eino/core_modules/components/) For more details about component implementations, please refer to the [Eino ecosystem documentation.](https://www.cloudwego.io/zh/docs/eino/ecosystem_integration/) -- **callback handlers**: official callback handlers implementing Eino's CallbackHandler interface, such as [Langfuse tracing](https://langfuse.com/docs/tracing) callback. - **DevOps tools**: IDE plugin for Eino that enables visualized debugging, UI based graph editing and more. For more details, please refer to the [Eino Dev tooling documentation.](https://www.cloudwego.io/zh/docs/eino/core_modules/devops/) ## Security diff --git a/README.zh_CN.md b/README.zh_CN.md index f6c6bc500..96e027590 100644 --- a/README.zh_CN.md +++ b/README.zh_CN.md @@ -8,24 +8,25 @@ EinoExt 项目为 [Eino](https://github.com/cloudwego/eino) 框架提供了各 - **组件实现**: Eino 组件类型的官方实现。 -| 组件类型 | 官方组件实现 | -|----------------------|----------------------------------------| -| ChatModel | OpenAI, Claude, Gemini, Ark, Ollama... | -| Tool | Google Search, Duck Duck Go... | -| Retriever | Elastic Search, Volc VikingDB... | -| ChatTemplate | DefaultChatTemplate... | -| Document Loader | WebURL, Amazon S3, File... | -| Document Transformer | HTMLSplitter, ScoreReranker... | -| Indexer | Elastic Search, Volc VikingDB... | -| Embedding | OpenAI, Ark... | -| Lambda | JSONMessageParser... | +### 组件概览 +| 组件类型 | 数量 | 官方实现 | +|---------|------|---------| +| **ChatModel** | 10 | OpenAI, Claude, Gemini, Ark, ArkBot, Ollama, Qwen, Qianfan, DeepSeek, OpenRouter | +| **Embedding** | 8 | OpenAI, Ark, Gemini, Ollama, Qianfan, DashScope, TencentCloud, Cache | +| **Indexer** | 10 | Elasticsearch (7/8/9), OpenSearch (2/3), Milvus, Milvus2, Qdrant, Redis, Volc VikingDB | +| **Retriever** | 12 | Elasticsearch (7/8/9), OpenSearch (2/3), Milvus, Milvus2, Qdrant, Redis, Dify, Volc VikingDB, Volc Knowledge | +| **Tool** | 10 | Bing Search, DuckDuckGo, Google Search, Wikipedia, SearXNG, BrowserUse, Command Line, HTTP Request, MCP, Sequential Thinking | +| **Prompt** | 2 | CozeLoop, MCP | +| **Document** | 12 | File/URL/S3 加载器, HTML/PDF/DOCX/XLSX 解析器, Recursive/HTML/Markdown/Semantic 分割器, Score Reranker | +| **Callback Handler** | 4 | APMPlus, CozeLoop, Langfuse, Langsmith | + +📋 **查看详细的组件信息（导入路径、描述、功能特性和 GitHub 链接），请参阅 [components.md](components.md)** 有关组件类型的更多详细信息，请参阅 [Eino 组件文档.](https://www.cloudwego.io/zh/docs/eino/core_modules/components/) 有关组件实现的更多详细信息，请参阅 [Eino 生态系统文档.](https://www.cloudwego.io/zh/docs/eino/ecosystem_integration/) -- **callback handlers**: 实现 Eino 的 callbacks.Handler 接口的官方 callback handler，例如[Langfuse tracing](https://langfuse.com/docs/tracing) 回调. - **DevOps 工具**: 用于 Eino 的 IDE 插件，支持可视化调试、基于 UI 的图形编辑等功能。更多详细信息，请参阅 [Eino Dev 工具文档.](https://www.cloudwego.io/zh/docs/eino/core_modules/devops/) ## 安全 diff --git a/components.md b/components.md new file mode 100644 index 000000000..2a51cc8c4 --- /dev/null +++ b/components.md @@ -0,0 +1,189 @@ +# Eino-Ext Components + +This document provides a comprehensive list of all component implementations in the Eino-Ext project, categorized by type. + +## Table of Contents + +- [ChatModel Components](#chatmodel-components) +- [Embedding Components](#embedding-components) +- [Indexer Components](#indexer-components) +- [Retriever Components](#retriever-components) +- [Tool Components](#tool-components) +- [Prompt Components](#prompt-components) +- [Document Components](#document-components) +- [Callback Handlers](#callback-handlers) + +--- + +## ChatModel Components + +ChatModel components provide integrations with various Large Language Model (LLM) providers for chat-based interactions. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| OpenAI | `github.com/cloudwego/eino-ext/components/model/openai` | OpenAI API integration for GPT models, providing access to GPT-4, GPT-3.5-turbo, and other OpenAI models. | • Support for GPT-4, GPT-3.5-turbo
• Streaming and non-streaming responses
• Function/tool calling
• Vision capabilities (GPT-4V)
• Configurable parameters
• Custom API base URL support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/openai) | +| Claude | `github.com/cloudwego/eino-ext/components/model/claude` | Anthropic Claude model integration for accessing Claude 3 family models. | • Support for Claude 3 family (Opus, Sonnet, Haiku)
• Streaming responses
• Tool use capabilities
• Vision support
• Long context windows (up to 200K tokens)
• System prompts support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/claude) | +| Gemini | `github.com/cloudwego/eino-ext/components/model/gemini` | Google Gemini model integration for Eino framework. | • Support for Gemini Pro and Gemini Pro Vision
• Streaming and non-streaming responses
• Tool/function calling support
• Multi-modal capabilities (text and images)
• Configurable temperature, top-p, top-k
• Safety settings configuration | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/gemini) | +| Ark | `github.com/cloudwego/eino-ext/components/model/ark` | ByteDance Ark platform model integration for accessing ByteDance's model platform. | • Access to ByteDance's model platform
• Streaming and non-streaming chat
• Tool calling support
• Multi-modal capabilities
• Image generation support
• Enterprise-grade reliability | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/ark) | +| ArkBot | `github.com/cloudwego/eino-ext/components/model/arkbot` | ByteDance ArkBot integration for conversational AI applications. | • Bot-specific API integration
• Streaming responses
• Context management
• Tool integration
• Conversation history handling | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/arkbot) | +| Ollama | `github.com/cloudwego/eino-ext/components/model/ollama` | Local Ollama model integration for running LLMs locally without API dependencies. | • Run models locally without API keys
• Support for various open-source models (Llama, Mistral, etc.)
• Streaming responses
• Tool calling support
• Customizable model parameters
• Multi-modal capabilities | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/ollama) | +| Qwen | `github.com/cloudwego/eino-ext/components/model/qwen` | Alibaba Qwen (Tongyi Qianwen) model integration. | • Support for multiple Qwen model variants
• Streaming and non-streaming chat
• Tool calling support
• Configurable generation parameters
• Multi-modal support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/qwen) | +| Qianfan | `github.com/cloudwego/eino-ext/components/model/qianfan` | Baidu Qianfan platform model integration for accessing Baidu's ERNIE models. | • Support for Baidu's ERNIE models
• Streaming responses
• Function calling
• Configurable API endpoints
• Authentication via API key and secret key | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/qianfan) | +| DeepSeek | `github.com/cloudwego/eino-ext/components/model/deepseek` | DeepSeek AI model integration for cost-effective inference. | • Support for DeepSeek-V2 and other variants
• Streaming chat completions
• Function calling
• Configurable generation parameters
• Cost-effective inference | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/deepseek) | +| OpenRouter | `github.com/cloudwego/eino-ext/components/model/openrouter` | OpenRouter API integration for accessing multiple LLM providers through a unified interface. | • Access to 100+ models from various providers
• Unified API interface
• Streaming support
• Tool calling capabilities
• Cost tracking and model routing | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/model/openrouter) | + +--- + +## Embedding Components + +Embedding components provide text embedding generation capabilities for semantic search and similarity tasks. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| OpenAI Embedding | `github.com/cloudwego/eino-ext/components/embedding/openai` | OpenAI embedding service integration for generating text embeddings using OpenAI's embedding models. | • Support for text-embedding-3-small, text-embedding-3-large, and ada-002
• Batch processing support
• High-quality embeddings
• Configurable dimensions | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/openai) | +| Ark Embedding | `github.com/cloudwego/eino-ext/components/embedding/ark` | ByteDance Ark platform embedding service integration. | • ByteDance embedding models
• Batch processing
• Enterprise support
• High performance | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/ark) | +| Gemini Embedding | `github.com/cloudwego/eino-ext/components/embedding/gemini` | Google Gemini embedding service integration. | • Google's embedding models
• Multi-language support
• Batch processing
• High-quality embeddings | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/gemini) | +| Ollama Embedding | `github.com/cloudwego/eino-ext/components/embedding/ollama` | Local embedding generation using Ollama for privacy-focused applications. | • Run embeddings locally
• No API key required
• Support for various embedding models
• Privacy-focused (data stays local)
• Batch embedding support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/ollama) | +| Qianfan Embedding | `github.com/cloudwego/eino-ext/components/embedding/qianfan` | Baidu Qianfan platform embedding service integration. | • Baidu embedding models
• Chinese language optimization
• Batch processing
• API-based service | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/qianfan) | +| DashScope Embedding | `github.com/cloudwego/eino-ext/components/embedding/dashscope` | Alibaba DashScope embedding service integration. | • Alibaba Cloud embedding models
• Multi-language support
• Batch processing
• Cloud-based service | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/dashscope) | +| TencentCloud Embedding | `github.com/cloudwego/eino-ext/components/embedding/tencentcloud` | Tencent Cloud embedding service integration. | • High-quality text embeddings
• Batch processing support
• Multiple embedding models
• Scalable cloud infrastructure
• Chinese language optimization | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/tencentcloud) | +| Cache Embedding | `github.com/cloudwego/eino-ext/components/embedding/cache` | Cache embedder for storing and retrieving embeddings efficiently to speed up the embedding process. | • Cache embeddings to avoid recomputation
• Support for different caching backends (Redis)
• Customizable key generation (hash-based)
• Transparent caching layer
• Performance optimization | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/embedding/cache) | + +--- + +## Indexer Components + +Indexer components provide vector database indexing capabilities for storing and managing embeddings. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| Elasticsearch 7 | `github.com/cloudwego/eino-ext/components/indexer/es7` | Elasticsearch 7.x indexer integration for full-text search capabilities. | • Elasticsearch 7.x support
• Traditional full-text search
• Aggregations
• Index management
• Query DSL | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/es7) | +| Elasticsearch 8 | `github.com/cloudwego/eino-ext/components/indexer/es8` | Elasticsearch 8.x indexer integration with vector search support. | • Elasticsearch 8.x compatibility
• Vector search support
• Full-text capabilities
• Index lifecycle management
• Security features | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/es8) | +| Elasticsearch 9 | `github.com/cloudwego/eino-ext/components/indexer/es9` | Elasticsearch 9.x indexer integration with latest features. | • Latest Elasticsearch features
• Dense vector search
• Sparse vector support
• Full-text search
• Advanced analytics
• Scalable indexing | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/es9) | +| OpenSearch 2 | `github.com/cloudwego/eino-ext/components/indexer/opensearch2` | OpenSearch 2.x indexer integration. | • OpenSearch 2.x support
• Vector search capabilities
• Full-text search
• Index management
• Query optimization | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/opensearch2) | +| OpenSearch 3 | `github.com/cloudwego/eino-ext/components/indexer/opensearch3` | OpenSearch 3.x indexer integration. | • Full-text and vector search
• OpenSearch 3.x compatibility
• Advanced query DSL
• Aggregations support
• Distributed search | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/opensearch3) | +| Milvus | `github.com/cloudwego/eino-ext/components/indexer/milvus` | Milvus vector database indexer integration. | • Scalable vector database
• Multiple index types (IVF, HNSW, etc.)
• Hybrid search support
• Distributed architecture
• High throughput | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/milvus) | +| Milvus2 | `github.com/cloudwego/eino-ext/components/indexer/milvus2` | Milvus 2.x version indexer integration with enhanced features. | • Updated Milvus 2.x API
• Improved performance
• Enhanced features
• Better scalability
• Collection management
• Multiple index building strategies (HNSW, Auto, Hybrid, etc.) | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/milvus2) | +| Qdrant | `github.com/cloudwego/eino-ext/components/indexer/qdrant` | Qdrant vector database indexer integration. | • High-performance vector storage
• HNSW algorithm for fast search
• Filtering and payload support
• Distributed deployment
• REST and gRPC APIs | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/qdrant) | +| Redis | `github.com/cloudwego/eino-ext/components/indexer/redis` | Redis vector database indexer integration. | • Redis-based vector storage
• Fast in-memory operations
• Vector similarity search
• Hybrid search capabilities
• Scalable architecture | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/redis) | +| Volc VikingDB | `github.com/cloudwego/eino-ext/components/indexer/volc_vikingdb` | ByteDance Volcengine VikingDB indexer integration. | • Volcengine cloud vector database
• High-performance indexing
• Scalable infrastructure
• Enterprise support
• Advanced filtering | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/indexer/volc_vikingdb) | + +--- + +## Retriever Components + +Retriever components provide document retrieval capabilities from various vector databases and knowledge bases. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| Elasticsearch 7 | `github.com/cloudwego/eino-ext/components/retriever/es7` | Elasticsearch 7.x retriever integration for document retrieval. | • ES 7.x support
• Traditional search
• Query DSL
• Aggregations
• Relevance tuning
• Multiple search modes (exact match, raw string) | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/es7) | +| Elasticsearch 8 | `github.com/cloudwego/eino-ext/components/retriever/es8` | Elasticsearch 8.x retriever integration with vector search. | • ES 8.x compatibility
• Vector search
• Full-text retrieval
• Query optimization
• Score boosting
• Dense vector similarity search
• Sparse vector search | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/es8) | +| Elasticsearch 9 | `github.com/cloudwego/eino-ext/components/retriever/es9` | Elasticsearch 9.x retriever integration with latest features. | • Latest ES features
• Dense vector retrieval
• Sparse vector retrieval
• Full-text search
• Hybrid retrieval
• Advanced ranking | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/es9) | +| OpenSearch 2 | `github.com/cloudwego/eino-ext/components/retriever/opensearch2` | OpenSearch 2.x retriever integration. | • OpenSearch 2.x compatibility
• Vector similarity search
• Full-text search
• Combined retrieval strategies
• Relevance scoring | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/opensearch2) | +| OpenSearch 3 | `github.com/cloudwego/eino-ext/components/retriever/opensearch3` | OpenSearch 3.x retriever integration. | • Vector and full-text retrieval
• Hybrid search
• Query DSL support
• Aggregation-based retrieval
• Score normalization | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/opensearch3) | +| Milvus | `github.com/cloudwego/eino-ext/components/retriever/milvus` | Milvus vector database retriever. | • High-performance vector search
• Multiple distance metrics
• Hybrid search support
• Filtering expressions
• Top-K retrieval | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/milvus) | +| Milvus2 | `github.com/cloudwego/eino-ext/components/retriever/milvus2` | Milvus 2.x retriever integration with advanced search modes. | • Milvus 2.x API support
• Enhanced search performance
• Advanced filtering
• Partition support
• Dynamic field filtering
• Multiple search modes (hybrid, iterator, range, scalar, sparse) | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/milvus2) | +| Qdrant | `github.com/cloudwego/eino-ext/components/retriever/qdrant` | Qdrant vector database retriever. | • Fast similarity search
• Filtering capabilities
• Payload-based retrieval
• Score threshold filtering
• Batch retrieval | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/qdrant) | +| Redis | `github.com/cloudwego/eino-ext/components/retriever/redis` | Redis vector database retriever. | • Fast in-memory retrieval
• Vector similarity search
• Hybrid search
• Filtering support
• High performance | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/redis) | +| Dify | `github.com/cloudwego/eino-ext/components/retriever/dify` | Dify platform retriever integration for knowledge base retrieval. | • Dify knowledge base integration
• API-based retrieval
• Multi-source retrieval
• Relevance scoring
• Context management | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/dify) | +| Volc VikingDB | `github.com/cloudwego/eino-ext/components/retriever/volc_vikingdb` | ByteDance Volcengine VikingDB retriever integration. | • Volcengine cloud vector database retrieval
• High-performance search
• Advanced filtering
• Enterprise support
• Scalable architecture | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/volc_vikingdb) | +| Volc Knowledge | `github.com/cloudwego/eino-ext/components/retriever/volc_knowledge` | ByteDance Volcengine Knowledge Base retriever integration. | • Volcengine knowledge base integration
• Semantic search
• Context-aware retrieval
• Metadata support
• Enterprise features | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/retriever/volc_knowledge) | + +--- + +## Tool Components + +Tool components provide various utilities and integrations for extending LLM capabilities. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| Bing Search | `github.com/cloudwego/eino-ext/components/tool/bingsearch` | Microsoft Bing search engine integration for web search capabilities. | • Bing Web Search API
• Rich search results
• Image and video search
• News search
• Entity recognition
• Market-specific results | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/bingsearch) | +| DuckDuckGo | `github.com/cloudwego/eino-ext/components/tool/duckduckgo` | DuckDuckGo search engine integration for privacy-focused web search. | • Privacy-focused search
• Web search capabilities
• No tracking
• Instant answers
• Safe search | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/duckduckgo) | +| Google Search | `github.com/cloudwego/eino-ext/components/tool/googlesearch` | Google Custom Search API integration. | • Google Custom Search API
• Programmable search
• Rich search results
• Customizable search parameters
• Safe search options | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/googlesearch) | +| Wikipedia | `github.com/cloudwego/eino-ext/components/tool/wikipedia` | Wikipedia search and content retrieval tool. | • Search Wikipedia articles
• Retrieve article content
• Summary extraction
• Multi-language support
• Structured data extraction | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/wikipedia) | +| SearXNG | `github.com/cloudwego/eino-ext/components/tool/searxng` | SearXNG metasearch engine integration for privacy-focused search. | • Privacy-focused search
• Multiple search engine aggregation
• Customizable search sources
• No tracking
• Self-hostable | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/searxng) | +| BrowserUse | `github.com/cloudwego/eino-ext/components/tool/browseruse` | Browser automation and web interaction tool for automated web browsing. | • Browser automation
• Web page interaction
• Element selection
• Screenshot capture
• Navigation control
• JavaScript execution | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/browseruse) | +| Command Line | `github.com/cloudwego/eino-ext/components/tool/commandline` | Execute command line operations with security constraints. | • Execute shell commands
• Python code execution
• File editing capabilities
• Capture stdout/stderr
• Working directory configuration
• Environment variables support
• Timeout control
• Security constraints | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/commandline) | +| HTTP Request | `github.com/cloudwego/eino-ext/components/tool/httprequest` | Generic HTTP request tool for API calls. | • Make HTTP requests (GET, POST, PUT, DELETE, etc.)
• Custom headers support
• Request body configuration
• Response parsing
• Timeout configuration
• Authentication support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/httprequest) | +| MCP (Model Context Protocol) | `github.com/cloudwego/eino-ext/components/tool/mcp` | Model Context Protocol tool integration for standardized tool communication. | • Standardized tool protocol
• Dynamic tool discovery
• Context management
• Tool chaining
• Protocol-based communication | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/mcp) | +| Sequential Thinking | `github.com/cloudwego/eino-ext/components/tool/sequentialthinking` | Tool for structured sequential reasoning and thinking processes. | • Step-by-step reasoning
• Thought chain management
• Structured thinking process
• Reasoning transparency
• Decision tracking | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/tool/sequentialthinking) | + +--- + +## Prompt Components + +Prompt components provide prompt management and template capabilities. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| CozeLoop Prompt | `github.com/cloudwego/eino-ext/components/prompt/cozeloop` | CozeLoop platform prompt integration for prompt template management. | • CozeLoop prompt templates
• Variable substitution
• Prompt optimization
• Template management
• Multi-language support | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/prompt/cozeloop) | +| MCP Prompt | `github.com/cloudwego/eino-ext/components/prompt/mcp` | Model Context Protocol prompt management. | • MCP-compliant prompts
• Dynamic prompt templates
• Context injection
• Prompt versioning
• Structured prompt format | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/prompt/mcp) | + +--- + +## Document Components + +Document components provide document loading, parsing, and transformation capabilities. + +### Document Loaders + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| File Loader | `github.com/cloudwego/eino-ext/components/document/loader/file` | Load documents from local file system. | • Load files from local filesystem
• Support for various file formats
• Batch loading
• Metadata extraction | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/loader/file) | +| URL Loader | `github.com/cloudwego/eino-ext/components/document/loader/url` | Load documents from web URLs. | • Load content from URLs
• HTTP/HTTPS support
• Authentication support
• Proxy configuration
• Custom headers
• Timeout configuration | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/loader/url) | +| S3 Loader | `github.com/cloudwego/eino-ext/components/document/loader/s3` | Load documents from Amazon S3 or S3-compatible storage. | • AWS S3 integration
• S3-compatible storage support
• Batch loading
• Credential management
• Prefix-based filtering | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/loader/s3) | + +### Document Parsers + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| HTML Parser | `github.com/cloudwego/eino-ext/components/document/parser/html` | Parse HTML documents and extract text content. | • HTML parsing
• Text extraction
• Tag filtering
• Structure preservation
• Metadata extraction | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/parser/html) | +| PDF Parser | `github.com/cloudwego/eino-ext/components/document/parser/pdf` | Parse PDF documents and extract text content. | • PDF text extraction
• Multi-page support
• Layout preservation
• Metadata extraction
• Image handling | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/parser/pdf) | +| DOCX Parser | `github.com/cloudwego/eino-ext/components/document/parser/docx` | Parse Microsoft Word (DOCX) documents. | • Parse DOCX files
• Extract text content
• Preserve formatting information
• Extract tables and images
• Metadata extraction
• Style preservation | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/parser/docx) | +| XLSX Parser | `github.com/cloudwego/eino-ext/components/document/parser/xlsx` | Parse Excel (XLSX) files for table data extraction. | • Support for Excel files with or without headers
• Select specific worksheets to process
• Custom document ID prefixes
• Automatic conversion of table data to document format
• Preservation of complete row data as metadata
• Support for additional metadata injection | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/parser/xlsx) | + +### Document Transformers + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| Recursive Splitter | `github.com/cloudwego/eino-ext/components/document/transformer/splitter/recursive` | Recursive text splitter for chunking long documents. | • Split text into chunks recursively
• Configurable chunk size
• Overlap size configuration to maintain context
• Useful for processing long documents
• Preserves context between chunks | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/transformer/splitter/recursive) | +| HTML Splitter | `github.com/cloudwego/eino-ext/components/document/transformer/splitter/html` | HTML-aware text splitter that respects HTML structure. | • HTML-aware splitting
• Preserve HTML structure
• Tag-based chunking
• Configurable chunk size
• Semantic splitting | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/transformer/splitter/html) | +| Markdown Splitter | `github.com/cloudwego/eino-ext/components/document/transformer/splitter/markdown` | Markdown-aware text splitter that respects Markdown structure. | • Markdown-aware splitting
• Header-based chunking
• Structure preservation
• Configurable chunk size
• Semantic splitting | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/transformer/splitter/markdown) | +| Semantic Splitter | `github.com/cloudwego/eino-ext/components/document/transformer/splitter/semantic` | Semantic text splitter that uses embeddings for intelligent chunking. | • Semantic-aware splitting
• Embedding-based chunking
• Intelligent boundary detection
• Context preservation
• Configurable similarity threshold | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/transformer/splitter/semantic) | +| Score Reranker | `github.com/cloudwego/eino-ext/components/document/transformer/reranker/score` | Score-based document reranker for improving retrieval results. | • Score-based reranking
• Relevance optimization
• Configurable scoring methods
• Result filtering
• Top-K selection | [Link](https://github.com/cloudwego/eino-ext/tree/main/components/document/transformer/reranker/score) | + +--- + +## Callback Handlers + +Callback handlers provide observability and tracing capabilities for Eino applications. + +| Name | Import Path | Description | Key Features | GitHub URL | +|------|-------------|-------------|--------------|------------| +| APMPlus | `github.com/cloudwego/eino-ext/callbacks/apmplus` | Volcengine APMPlus callback implementation for enhanced observability. | • Implements Eino Handler interface
• Session functionality for request association
• Easy integration with Eino applications
• Trace and metrics reporting
• Enterprise monitoring | [Link](https://github.com/cloudwego/eino-ext/tree/main/callbacks/apmplus) | +| CozeLoop | `github.com/cloudwego/eino-ext/callbacks/cozeloop` | CozeLoop callback implementation for Eino observability. | • Implements Eino Handler interface
• Easy integration with Eino applications
• CozeLoop platform integration
• Trace collection
• Performance monitoring | [Link](https://github.com/cloudwego/eino-ext/tree/main/callbacks/cozeloop) | +| Langfuse | `github.com/cloudwego/eino-ext/callbacks/langfuse` | Langfuse tracing callback for LLM observability. | • Implements Eino Handler interface
• Langfuse platform integration
• Trace collection
• Cost tracking
• Performance analytics | [Link](https://github.com/cloudwego/eino-ext/tree/main/callbacks/langfuse) | +| Langsmith | `github.com/cloudwego/eino-ext/callbacks/langsmith` | Langsmith tracing callback for LLM observability and debugging. | • Implements Eino Handler interface
• Langsmith platform integration
• Trace collection
• Session management
• Debugging support | [Link](https://github.com/cloudwego/eino-ext/tree/main/callbacks/langsmith) | + +--- + +## Summary + +This document cataloged **64 component implementations** across the following categories: + +- **10 ChatModel** implementations (OpenAI, Claude, Gemini, Ark, ArkBot, Ollama, Qwen, Qianfan, DeepSeek, OpenRouter) +- **8 Embedding** implementations (OpenAI, Ark, Gemini, Ollama, Qianfan, DashScope, TencentCloud, Cache) +- **10 Indexer** implementations (ES7, ES8, ES9, OpenSearch2, OpenSearch3, Milvus, Milvus2, Qdrant, Redis, Volc VikingDB) +- **12 Retriever** implementations (ES7, ES8, ES9, OpenSearch2, OpenSearch3, Milvus, Milvus2, Qdrant, Redis, Dify, Volc VikingDB, Volc Knowledge) +- **10 Tool** implementations (Bing Search, DuckDuckGo, Google Search, Wikipedia, SearXNG, BrowserUse, Command Line, HTTP Request, MCP, Sequential Thinking) +- **2 Prompt** implementations (CozeLoop, MCP) +- **12 Document** implementations (3 loaders, 4 parsers, 5 transformers) +- **4 Callback** handlers (APMPlus, CozeLoop, Langfuse, Langsmith) + +All components are designed to work seamlessly within the [Eino framework](https://github.com/cloudwego/eino) ecosystem and provide standardized interfaces for their respective functionalities. + +--- + +**Project Repository:** https://github.com/cloudwego/eino-ext + +**Documentation:** https://www.cloudwego.io/docs/eino/