wet

WET is an MCP server that aggregates web search, content extraction, and document/documentation indexing with multimodal capabilities. It leverages an embedded SearXNG instance for web search, automatic content extraction (Markdown/Text), and a local cache to accelerate repeated queries. The server can discover and index library docs, perform deep crawls from root URLs, and support media listing and downloads. Additionally, it includes optional cloud embedding and reranking capabilities, and can sync indexed docs across machines via rclone. Use the provided tools to search the web, fetch relevant documents, extract clean content, and build searchable indexes for local or remote usage.

Prerequisites:

• Python 3.13 installed (3.14+ is not supported due to SearXNG compatibility)
• Optional: uvx (Python-based runtime) or Docker if you prefer containerized deployment

Install using uvx (recommended):

# Ensure Python 3.13 is available and uvx is installed
# If you don't have uvx installed, install via your preferred method (see project docs)

# Start the Wet MCP server using uvx with Python 3.13
uvx --python 3.13 wet-mcp@latest

Install using Docker (alternative):

# Pull and run the Wet MCP image with a persistent data volume
docker run -i --rm \
  --name mcp-wet \
  -v wet-data:/data \
  -e API_KEYS="GOOGLE_API_KEY:AIza..." \
  -e SYNC_ENABLED="true" \
  -e SYNC_REMOTE="gdrive" \
  -e RCLONE_CONFIG_GDRIVE_TYPE="drive" \
  -e RCLONE_CONFIG_GDRIVE_TOKEN="<base64>" \
  n24q02m/wet-mcp:latest

Pre-install (optional):

# Pre-download SearXNG, Playwright, and embedding/reranker models to speed up first run
uvx --python 3.13 wet-mcp warmup

# If using cloud embedding, you can preset API keys during warmup
API_KEYS="GOOGLE_API_KEY:AIza..." uvx --python 3.13 wet-mcp warmup

Sync setup (one-time):

# Google Drive
uvx --python 3.13 wet-mcp setup-sync drive

# Other providers (example: Dropbox, OneDrive, S3)
uvx --python 3.13 wet-mcp setup-sync dropbox
uvx --python 3.13 wet-mcp setup-sync onedrive
uvx --python 3.13 wet-mcp setup-sync s3

Notes:

• The first run may download substantial assets (SearXNG, Playwright binaries, and embedding/reranker models). Ensure adequate disk space and network access.
• If you enable SYNC_ENABLED, configure the remote storage (e.g., Google Drive) and provide the appropriate RCLONE tokens as described in the docs.
• For cloud embedding and media analysis, set API_KEYS with your Google/embedding provider keys; otherwise, Wet uses local ONNX-based embeddings and reranker models.
• When using uvx, you must specify the exact Python version (3.13) to avoid compatibility issues with SearXNG.
• The embedded search index supports rapid local searches; for larger deployments, consider enabling SYNC to keep docs synchronized across machines.