hayhooks

Overview

Hayhooks is a versatile tool that allows developers to deploy Haystack pipelines and agents as both REST APIs and MCP tools. It provides seamless integration with AI development environments like Cursor and Claude Desktop, enabling pipelines to be accessible as tools. The project supports multiple deployment options including OpenAI-compatible endpoints, Chainlit UI integration, and Open WebUI compatibility. With features like automatic OpenAI endpoint generation, file upload support, and OpenTelemetry tracing, Hayhooks offers a comprehensive solution for serving machine learning pipelines.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• my_agent_run

  Run a deployed Haystack agent with a question

  Try:What can you do?Try:What's the weather in London?
• chat_completion

  Streaming chat completion endpoint for Haystack agents

  Try:Generate a response to the user's messageTry:Continue our conversation
• pipeline_deploy

  Deploy a Haystack pipeline to Hayhooks

  Try:Deploy my document processing pipelineTry:Set up my RAG system
• pipeline_undeploy

  Undeploy a Haystack pipeline from Hayhooks

  Try:Remove my deployed pipelineTry:Stop my chatbot service
• pipeline_list

  List all deployed Haystack pipelines

  Try:Show me all deployed pipelinesTry:What agents are running?

Note: Tool names inferred from the documentation about Hayhooks exposing Haystack pipelines as MCP tools, with tool names based on the pipeline naming pattern and API endpoints mentioned in the README.

Comparable tools

Installation

# Install Hayhooks
pip install hayhooks

# Start Hayhooks
hayhooks run

# Deploy a pipeline
hayhooks pipeline deploy-files -n my_agent ./my_agent_dir

{
  "mcpServers": {
    "hayhooks": {
      "command": "python",
      "args": ["-m", "hayhooks", "run"],
      "env": {}
    }
  }
}

FAQ

What is the difference between Hayhooks and other Haystack deployment tools?

Hayhooks specifically focuses on exposing Haystack pipelines as both REST APIs and MCP tools, making it ideal for AI development environments. It provides built-in MCP server functionality with minimal configuration.

Can I integrate Hayhooks with existing AI development environments?

Yes, Hayhooks is designed to work seamlessly with environments like Cursor and Claude Desktop through the MCP protocol. It can also be integrated with Open WebUI and Chainlit.

Compare hayhooks with