MCP.so
登录

Alation AI Agent SDK

@Alation

关于 Alation AI Agent SDK

Python library for integrating the Alation API into agentic workflows

基本信息

分类

开发工具

许可证

Apache-2.0

运行时

python

传输方式

stdio

发布者

Alation

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "alation-ai-agent-sdk": {
      "command": "uv",
      "args": [
        "pip",
        "install",
        "alation-ai-agent-sdk==1.0.0rc3"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is Alation AI Agent SDK?

The Alation AI Agent SDK enables AI agents to access and leverage metadata from the Alation Data Catalog. It is designed for developers and data teams who want to integrate AI agents with their data catalog for tasks like search, discovery, and curation.

How to use Alation AI Agent SDK?

Install the SDK via pip with uv, then configure it with your Alation instance credentials using ServiceAccountAuthParams. You can use the core SDK directly, or integrate with LangChain or MCP (Model Context Protocol) via separate packages.

Key features of Alation AI Agent SDK

  • Integrates with Alation Data Catalog for metadata access
  • Supports natural language search for relevant metadata
  • Includes MCP and LangChain integrations
  • Provides tools like catalog context search, lineage, and data quality
  • Supports streaming and chat ID for enhanced agent interactions
  • Offers a remote MCP server option for easier deployment

Use cases of Alation AI Agent SDK

  • Asset curation and metadata enrichment
  • Search and discovery across catalog assets
  • Building role-based and data analyst agents
  • Creating custom query flow agents
  • Integrating with AI frameworks like MCP and LangChain

FAQ from Alation AI Agent SDK

What authentication methods are supported?

The recommended method is Service Account authentication using client_id and client_secret. User account authentication is documented as a fallback.

What are the runtime requirements?

Python 3.10 or higher, access to an Alation Data Catalog instance, and a valid refresh token or client credentials.

How does the remote MCP server work?

The remote MCP server allows clients to connect without installing the SDK locally. After a one-time authorization setup, you can add a URL like https://<your_instance>/ai/mcp to your MCP client.

What tools are included in the SDK?

Supported tools include alation_context, bulk_retrieval, data_quality_tool, get_custom_fields_definitions, get_data_products, get_data_sources_tool, get_signature_creation_instructions, and lineage.

What agents are available?

Agents include catalog_context_search_agent, custom_agent, query_flow_agent, and sql_query_agent.

评论

开发工具 分类下的更多 MCP 服务器