Web Search MCP Server with ChromaDB Vector Database

What is Web Search MCP Server with ChromaDB Vector Database?

This MCP server provides tools for web search and vector database functionality using LangChain and ChromaDB. It enables searching documentation for popular libraries (LangChain, LlamaIndex, OpenAI) and storing/retrieving documents via vector embeddings with semantic similarity search.

How to use Web Search MCP Server with ChromaDB Vector Database?

Install dependencies with pip install -e . (or uv pip install -e .), create a .env file with a Serper API key and ChromaDB settings, then run python main.py. The server exposes tools for web search and ChromaDB operations, which can be invoked programmatically.

Key features of Web Search MCP Server with ChromaDB Vector Database

• Search documentation for LangChain, LlamaIndex, and OpenAI
• Extract content from web pages
• Store documents with vector embeddings via ChromaDB
• Perform semantic similarity search with metadata filtering
• Batch add multiple documents for efficiency
• Create retrievers for downstream AI applications

Use cases of Web Search MCP Server with ChromaDB Vector Database

• Searching official documentation for library-specific queries
• Building a semantic search over a custom document collection
• Adding, querying, updating, and deleting documents programmatically
• Creating a retriever to feed relevant context to language models

FAQ from Web Search MCP Server with ChromaDB Vector Database

What dependencies does this server require?

Python packages listed in the project plus a Serper API key for web search and a sentence‑transformer embedding model.

How are documents persisted?

ChromaDB persists data to a directory specified by the CHROMA_PERSIST_DIRECTORY environment variable (default ./chroma_db).

What transport modes are supported?

The server supports both stdio and sse transport modes, configured via the TRANSPORT environment variable.

What is the Serper API used for?

The Serper API performs web searches; it is required and configured with SERPER_API_KEY and SERPER_API_URL.

Can I add multiple documents at once?

Yes, the batch_add_documents_to_vectordb tool accepts a list of documents for efficient batch insertion.