Document Cleanup Agent

Turn messy documents into clean, structured output. The dad test: if he can use it, anyone can.

Status: MVP in progress

Phase 1 Goal: Prove the engine works

• Input: Messy document (meeting notes, contracts, research, etc.)
• Output: Cleaned, structured document (markdown or JSON)

Quick Start

# Clean up a document from stdin
cat messy-notes.txt | python cleanup.py

# From a file
python cleanup.py --file meeting-notes.txt

# Output as JSON
python cleanup.py --file notes.txt --format json

What It Does

1. Extract text from various formats (txt, md, pdf, docx, html)
2. Identify document structure (headings, lists, paragraphs)
3. Fix formatting issues (spacing, bullets, numbering)
4. Generate clean, consistent output
5. Optionally extract metadata (dates, names, action items)

Examples

Input: Rambling meeting notes with inconsistent formatting Output: Structured summary with attendees, decisions, action items

Input: Scanned contract with OCR errors
Output: Clean text with sections properly identified

Stack

• Python 3.11+
• google-genai (Gemini API)
• python-docx (Word docs)
• PyMuPDF (PDFs)
• beautifulsoup4 (HTML)

Part of Reify Studio

This is one tool in the Reify Studio collection — AI tools that feed into your personal knowledge vault.