PDF Search for Zed

A document search extension for Zed that lets you semantically search through a PDF document and use the results in Zed's AI Assistant.

Prerequisites

This extension currently requires:

1. An OpenAI API key (to generate embeddings)
2. uv installed on your system

Note: While the current setup requires an OpenAI API key for generating embeddings, we plan to implement a self-contained alternative in future versions. Community feedback will help prioritize these improvements.

Quick Start

1. Clone the repository

git clone https://github.com/freespirit/pdfsearch-zed.git

1. Set up the Python environment for the MCP server:

cd pdfsearch-zed/pdf_rag
uv venv
uv sync

1. Install Dev Extension in Zed

2. Build the search db

cd /path/to/pdfsearch-zed/pdf_rag

echo "OPENAI_API_KEY=sk-..." > src/pdf_rag/.env

# This may take a couple of minutes, depending on the documents' size
# You can provide multiple files and directories as arguments.
#  - files would be chunked.
#  - a directory would be considered as if its files contains chunks.
#    E.g. they won't be further split.
uv run src/pdf_rag/rag.py build "file1.pdf" "dir1" "file2.md" ...

1. Configure Zed

"context_servers": {
    "pdfsearch-context-server": {
        "settings": {
            "extension_path": "/path/to/pdfsearch-zed"
        }
    }
}

Usage

1. Open Zed's AI Assistant panel
2. Type /pdfsearch followed by your search query
3. The extension will search the PDF and add relevant sections to the AI Assistant's context

Future Improvements

• Self-contained vector store
• Self-contained embeddings
• Automated index building on first run
• Configurable result size
• Support for multiple PDFs
• Optional: Additional file formats beyond PDF

Project Structure

• pdf_rag/: Python-based MCP server implementation
• src/: Zed extension code
• extension.toml and Cargo.toml: Zed extension configuration files

Known Limitations

• Manual index building is required before first use
• Requires external services (OpenAI)