Top 200 ReposEXPLORER
Browse agents, categories and pages
Back to results
Planned, not doneThe source describes an intention, not a finished result.

Convert PDFs and draft structured reports from a desktop assistant with Docling MCP

Docling MCP, from the IBM-started Docling project, gives Claude Desktop, LM Studio and other MCP clients tools to convert PDFs into structured documents and write new ones section by section.

The problem

How it works

Docling MCP exposes the Docling document library as MCP tools. You add one entry to Claude Desktop’s config or install it into LM Studio with a button, and it runs through uvx. In local mode it converts documents on your own machine; in remote mode it calls a Docling Serve instance. Conversion tools turn a PDF from a file path or URL into a structured DoclingDocument with a key the assistant can refer to later. Generation tools let the assistant create a new document, add a title, headings, paragraphs and lists step by step, check progress by exporting it to Markdown, and save the result. Optional tool groups add RAG with Milvus or LlamaIndex.

Results and limitations

The README gives setup steps and example prompts but no measured results, so this case is marked planned.

Before you try it

Local mode keeps documents on your machine, but if the assistant itself is a cloud model, the converted text it reads still goes to that provider; pair it with a local model in LM Studio for fully local work. Turning on picture descriptions adds a vision-model step for every figure and can make conversion much slower on a laptop, so raise the client’s timeout. Check converted tables against the original PDF. Results depend on your setup and were not independently reproduced by Top 200 Repos.

Tools and repositories

No ranked repository is involved. This use case still stands on its own source.

Tools
Docling MCP · uv · Claude Desktop or LM Studio
Prerequisites
Python with uv, and an MCP client; local conversion needs the docling-mcp[local] extra.

Try it yourself: Turn a set of PDFs into one structured brief

Fill the fields to adapt the prompt. Empty fields stay as visible [placeholders].

Preview
Convert each of these PDFs with the Docling tools: [PDF files]. Then write a new Docling document titled [Report title] for [Audience]. Use one section per [What each section covers], and under each section summarise the main points and copy any key figures exactly, naming the source file and page. Export the draft to Markdown so I can review it before you save the final document to [Output folder].

Compare every copied figure with the source PDF before sharing the report.

Sources

  1. docling-project/docling-mcp (opens in a new tab)

    Project documentation · github.com · Undated · Source checked 12 Nov 2026, 09:56 UTC

    What this source supports
    • Docling MCP provides tools for PDF conversion into DoclingDocument and for generating new documents
    • Local, remote (Docling Serve) and hybrid conversion modes
    • Setup for Claude for Desktop and LM Studio through uvx
    • Example prompts for converting a PDF and for writing a document with create, add title, headings, paragraphs and Markdown export tools
    • Picture description adds a VLM step per figure and can slow conversion; MIT licence; started by IBM Research Zurich, hosted by LF AI & Data
Report a problem

Reported by source

Convert PDFs and Office files to Markdown for an LLM

Run MarkItDown on PDFs, Word, PowerPoint and Excel files to get structure-preserving Markdown an LLM can read cheaply.

github.commicrosoft/markitdown
Reported by source

Parse scanned PDFs with tables into Markdown and JSON

Run PaddleOCR's PP-StructureV3 pipeline on scans to recover layout, tables, formulas and reading order as Markdown or JSON.

github.comPaddlePaddle/PaddleOCR
Reported by source

Extract invoice fields locally with a vision model on Ollama

Stephen Sebastian replaced a paid cloud OCR API with Gemma 4 E4B on Ollama to pull date, amount and description from contractor invoice photos on his laptop.

dev.toollama/ollama
Reported by source

Give Claude your Google Calendar and Gmail through a self-run MCP server

The open-source Google Workspace MCP server exposes Calendar, Gmail and other Workspace tools to Claude Desktop, Claude Code and other MCP clients, using your own OAuth client and optional read-only mode.

github.com
Reported by source

Let an assistant test QA tickets in a browser via Playwright MCP

Peter Eysermans connected Claude Desktop to Playwright and GitHub MCP servers so it could walk through features from tickets, report results and draft reusable tests.

madewithlove.com · 22 Jul 2025microsoft/playwright
Reported by source

Program a robot dog with an AI assistant: Anthropic's Project Fetch experiment

Anthropic split eight staff without robotics experience into two teams to program a quadruped robot to fetch a ball; the team using Claude finished more tasks and, on shared tasks, took about half the time.

anthropic.com · 12 Nov 2025