# docshelf-mcp

Source: https://glimind.com/server/pypi%2Fdocshelf-mcp
Registry id: `pypi/docshelf-mcp`
Status: unknown

## Install

```json
{
  "mcpServers": {
    "docshelf-mcp": {
      "command": "uvx",
      "args": [
        "docshelf-mcp"
      ]
    }
  }
}
```

Claude Code:

```sh
claude mcp add docshelf-mcp -- uvx docshelf-mcp
```

## Tools (11)

- `docshelf_add_directory`: Add every supported file in a folder in one call. Re-indexes once.
- `docshelf_add_document`: Add a file (MD/PDF/DOCX/HTML/EPUB). Converts, splits, re-indexes.
- `docshelf_convert_pdf`: Standalone PDF → Markdown (no shelf).
- `docshelf_doctor`: Check shelf integrity; optionally auto-fix safe drift.
- `docshelf_init_shelf`: Bootstrap a new shelf directory.
- `docshelf_list_documents`: List documents by category.
- `docshelf_read_document`: Read a document/section's content over MCP (works on private shelves).
- `docshelf_rebuild_index`: Regenerate INDEX.md from disk.
- `docshelf_remove_document`: Remove a document, its sections, and metadata. Re-indexes.
- `docshelf_rename_document`: Retitle / recategorize a document (moves file, sections, meta) — no re-conversion.
- `docshelf_search`: Plain-text search across the shelf, with raw URLs.

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.