MCPIndependentxberg-io
Xberg
Extracts text, tables and metadata from PDFs, Office files, scans and archives.
The Xberg engine (the successor to Kreuzberg), started with the xberg mcp command, detects the file type and extracts clean content, tables and metadata from PDFs, Office documents, scans (OCR), web pages and archives, one at a time or in batches, with caching. It runs locally on the CPU with no GPU needed; audio transcription requires an extra feature.
Installation
Manual install
$claude mcp add xberg -- npx -y @xberg-io/xberg-cli@latest mcp --transport stdioDetails
- Works in
- Claude, Claude Code, Cursor
- Transport
- local
- License
- MIT
- Tasks
- Docs, Data
- Rating
- Source
- github.com/xberg-io/xberg (opens in a new tab)
Tags
- extraction
- ocr
- tables
- metadata
- office
A pack is your own set of skills and MCP servers. sciver loads it in your AI with one command.