docx extraction

Turn Word documents into Markdown your agent can review.

Contracts, briefs, policies, and reports land as DOCX all the time. Frenchie gives your agent Markdown instead of asking it to guess through a binary file.

Install the skill pack Read docs →

$Extract ./contract-draft.docx to Markdown with Frenchie

✓ extracted · saved → result.md

# Extracted Markdown

## Sheet: Summary

| Field | Value |
| ----- | ----- |
| ARR   | $4.9M |
| Risk  | Procurement delay |

## Notes

- Source file preserved as
  agent-readable Markdown.
- Your agent does the reasoning
  after extraction.

why it exists

Files first. Reasoning after.

DOCX files hide useful structure behind an Office container. If your agent only sees an attachment, it may miss headings, lists, tables, or links that matter to the review.

Frenchie maps headings, paragraphs, lists, tables, hyperlinks, footnotes, and embedded images into Markdown references your agent can inspect.

quick start

Install skills and MCP tools together.

The installer configures the MCP server and installs the Frenchie skills alongside the MCP tools, so your agent has the callable tools and the instructions for when to use them.

Install Frenchie

$ npx @lab94/frenchie install --api-key fr_...

This writes the local stdio MCP config and copies the Frenchie skills into your project.

Ask your agent to extract

> Extract ./contract-draft.docx to Markdown with Frenchie

Your agent calls extract_to_markdown. Stdio results auto-save to .frenchie/<name>/result.md.

Reason over Markdown

DOCX extraction costs 0.5 credit per page, with a 0.5 credit minimum. DOCX files can be up to 50 MB. Results expire 30 minutes after first delivery.

workflow

How it runs.

1
Install Frenchie with the skill pack and MCP tools together.
2
Ask your agent to extract the DOCX file.
3
Frenchie turns the document structure into Markdown and saves the result.
4
Your agent reviews clauses, dates, tables, or edits from the Markdown it can actually read.

questions

The practical bits.

Does Frenchie read legacy .doc files?

No. Save the file as modern `.docx` first. Legacy binary Office formats are intentionally out of scope.

Does it keep comments and tracked changes?

No. Frenchie focuses on the readable document content. Comments and tracked changes are dropped in v1.

Can my agent extract specific fields from the Word document?

Yes, after conversion. Frenchie returns Markdown; your agent can then pull dates, clauses, names, or fields with its own prompt.

Give your agent a real file.

Install Frenchie, drop in a DOCX, XLSX, CSV, TSV, or PPTX file, and see the Markdown your agent gets back.

Install the skill pack Browse use cases →