HexDocs FAQs

Question 1

What is HexDocs MCP?

Accepted Answer

HexDocs MCP provides semantic search capabilities for Hex package documentation, designed for AI applications. It includes an Elixir package for processing documentation and a TypeScript server implementing the Model Context Protocol (MCP).

Question 2

How do I install HexDocs MCP?

Accepted Answer

The installation involves setting up both the Elixir package and the TypeScript MCP server. Refer to the 'Installation' section in the documentation for detailed steps, including cloning the repository, building the TypeScript server, and configuring your MCP client.

Question 3

What are the requirements for using HexDocs MCP?

Accepted Answer

HexDocs requires Ollama (with the 'nomic-embed-text' model pulled), Elixir 1.16+, and Node.js 22 or later. Ensure Ollama is running before using the embedding features.

Question 4

How do I use the Elixir package?

Accepted Answer

Use the `mix hex.docs.mcp fetch` command to download, process, and generate embeddings for a Hex package. You can also specify a version or a specific embedding model. The `mix hex.docs.mcp search` command allows you to search within existing embeddings.

Question 5

How do I configure the storage location for HexDocs MCP?

Accepted Answer

By default, data is stored in `~/.hexdocs_mcp`. You can change this location by setting the `HEXDOCS_MCP_PATH` environment variable, either in your shell or in the MCP server configuration.

HexDocs

About

Key Features

Use Cases

HexDocs

About

Key Features

Use Cases