doc_to_markdown
Convert images, PDFs, and Word documents to Markdown.
LABSCO SUMMARY
OCR narrowed to three calls, one of which answers a question instead of returning the whole page.
The distinction that matters is between converting a document to Markdown, which gives a model everything, and extraction, which takes a key or a table_header and returns only that — the difference between putting a page in the context window and putting a field there. If you refer to the extraction tool by name in a prompt, note that it is spelled general_information_extration.
What it is
A document-recognition server over TextIn, three tools that each take a local path: text recognition, targeted extraction, and Markdown conversion.
What you get
- Text recognised out of images, Word documents and PDF files
- Named values pulled from a document by key, and tables pulled out by their header row
- The same three formats converted to Markdown, structure intact, for a model to read
Requirements
APP_ID and APP_SECRET for a TextIn account, with MCP_SERVER_REQUEST_TIMEOUT alongside them in the same configuration. Every tool takes a path, so the documents have to sit where the server can open them.
Setup effort
One command plus credentials — npx -y @intsig/server-textin, then supply credentials
USE CASES
01
Reading a scanned document that has no text layer
Point recognition at the file and get its text back, for the photographed and scanned pages that copy-paste cannot touch.
02
Turning a report into something a knowledge base can hold
Convert the document to Markdown so headings and tables survive the trip, rather than flattening the whole thing into a wall of recognised text.
03
Pulling the three fields you actually need off a form
Name the keys you want and the header row of the table you want, and get those values back instead of the whole document — the difference between reading an invoice and processing a stack of them.
3 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
general_information_extration
Automatically identify and extract information from documents, or identify and extract user-specified information.
recognition_text
Text recognition from images, Word documents, and PDF files.
BEFORE YOU INSTALL
What to know first
Tools you get
3 tools when we connected.
Requires
- npx on PATH
- APP_ID in your environment
- APP_SECRET in your environment
CONTEXT COST
Light Standing cost
Rated by the size of the 3 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Add the block to your client, or run the command directly.
Server setup
Configuration file
JSON
{
"mcpServers": {
"textin-ocr": {
"command": "npx",
"args": [
"-y",
"@intsig/server-textin"
],
"env": {
"APP_ID": "<YOUR_APP_ID>",
"APP_SECRET": "<YOUR_APP_SECRET>",
"MCP_SERVER_REQUEST_TIMEOUT": "600000"
},
"timeout": 600
}
}
}In the Config file block, <YOUR_APP_ID> and <YOUR_APP_SECRET> are the APP_ID and APP_SECRET of your TextIn account — TextIn's guide explains where to find them: https://www.textin.com/doc/guide/account/%E5%A6%82%E4%BD%95%E8%8E%B7%E5%8F%96app%20id?status=first. The npx command has no block: run it with APP_ID and APP_SECRET set in your environment.
npx
TERMINAL
npx -y @intsig/server-textin
