hayhooks
by deepset-ai
Overview
Deploy and serve Haystack AI pipelines and agents as REST APIs or MCP tools, enabling integration with Open WebUI and other OpenAI-compatible clients for conversational AI, RAG systems, and custom AI applications.
Installation
hayhooks runEnvironment Variables
- OPENAI_API_KEY
- HUGGING_FACE_API_KEY
- HAYHOOKS_HOST
- HAYHOOKS_PORT
- HAYHOOKS_MCP_HOST
- HAYHOOKS_MCP_PORT
- HAYHOOKS_PIPELINES_DIR
- HAYHOOKS_ADDITIONAL_PYTHON_PATH
- LOG
- HAYHOOKS_USE_HTTPS
- HAYHOOKS_DISABLE_SSL
- HAYHOOKS_SHOW_TRACEBACKS
- HAYHOOKS_STREAMING_COMPONENTS
- HAYHOOKS_CORS_ALLOW_ORIGINS
- HAYHOOKS_CORS_ALLOW_METHODS
- HAYHOOKS_CORS_ALLOW_HEADERS
- HAYHOOKS_CORS_ALLOW_CREDENTIALS
- HAYHOOKS_CORS_ALLOW_ORIGIN_REGEX
- HAYHOOKS_CORS_EXPOSE_HEADERS
- HAYHOOKS_CORS_MAX_AGE
Security Notes
The server's core functionality allows deploying and executing user-provided Python code (e.g., `pipeline_wrapper.py`) via API endpoints. This introduces a significant Remote Code Execution (RCE) risk if deployment endpoints are not strictly secured by external access control mechanisms. The default CORS settings are overly permissive (`*`) for production environments, and the server does not include built-in authentication, requiring external layers for security. While `show_tracebacks` is false by default, preventing some information leakage, the potential for arbitrary code execution mandates careful deployment in trusted environments only.
Similar Servers
mcpo
Exposes Model Context Protocol (MCP) tools as OpenAPI-compatible HTTP servers for integration with LLM agents and other applications.
mcp-openapi-server
Exposes OpenAPI endpoints as Model Context Protocol (MCP) tools, enabling Large Language Models (LLMs) to discover and interact with REST APIs through a standardized protocol.
mcp-servers
Provides a curated collection of Model Context Protocol (MCP) server configurations to enable AI agents to interact with various developer tools and services.
zeromcp
A minimal, pure Python Model Context Protocol (MCP) server for exposing tools, resources, and prompts via HTTP/SSE and Stdio transports.