WI Docs

A universal document toolkit — beyond PDF.

WI Docs is envisioned as a web-based workspace for working with document files across formats. Inspired by the convenience of iLovePDF, it aims to bring conversion, editing, organization and document utilities together in one place—not just for PDF, but for a broad range of document formats.

Project status: The current codebase is an early document-to-output application. The broader multi-format toolkit described here is the product direction, not a claim that every planned tool is already implemented.

Current implementation

The repository currently contains a React interface and an Express API focused on document composition and export.

• Import Markdown (.md), plain text (.txt) and Word Open XML (.docx) files.
• Edit content in a text editor and preview the rendered document.
• Convert DOCX input to HTML using Mammoth.
• Customize page size, margins, page/content colors, typography and background imagery.
• Generate PDF output using Puppeteer.
• Generate HTML output.
• Generate legacy Word-compatible .doc output from HTML.
• Undo and redo recent content changes in the editor.

These capabilities describe the inspected code paths; they should not be interpreted as a guarantee of production readiness or complete format fidelity.

Product vision

WI Docs aims to become a unified collection of document tools, including:

Convert

Convert supported document formats between PDF, DOCX, DOC, TXT, Markdown, RTF, ODT and other formats as compatible engines are introduced.

Organize

Merge, split, reorder, extract, rotate and rearrange pages or document sections where the format supports those operations.

Optimize

Compress files, optimize embedded images, reduce document size and prepare files for sharing.

Edit and prepare

Edit document content, adjust layouts, apply templates, add headers/footers, page numbers, watermarks and branding.

Extract and transform

Extract text and images, convert scans through OCR, and transform content into reusable formats.

Batch workflows

Process multiple files in one operation, with clear per-file status and downloadable results.

The supported format matrix and tool availability must be documented as each capability is implemented and tested.

Technology stack

Area	Current technology
Frontend	React 18, JavaScript, Vite 6
Styling	Tailwind CSS 3
Backend	Node.js, Express 4
PDF rendering	Puppeteer
Markdown	markdown-it
DOCX import	Mammoth
HTTP middleware	CORS

The client and API have separate package manifests: client/package.json and root package.json.

Repository layout

.
├── client/                 # React + Vite application
│   └── src/
│       ├── components/     # Editor, upload, preview and UI components
│       └── utils/          # Document conversion utilities
├── server.js               # Express API and document rendering endpoints
├── package.json            # Backend dependencies and scripts
├── DOCUMENTACION.md        # Legacy documentation (not authoritative)
└── docs/                   # Product and engineering documentation

Development setup

Requirements

• Node.js compatible with the installed dependencies.
• npm.
• Chromium dependencies required by Puppeteer on your operating system.

Install dependencies

From the repository root:

npm install
cd client
npm install

Run the API

In the repository root:

npm run dev

The API listens on port 3000 by default. Set PORT to override it.

Run the frontend

In a second terminal:

cd client
npm run dev

Vite serves the frontend. The API proxy configuration and deployment routing should be checked for the active environment before connecting the client to a separately hosted API.

Production build

cd client
npm run build

The current root package exposes start and dev scripts for the Express API. There is no root-level orchestrated build script in the inspected package manifest.

API endpoints currently present

Method	Endpoint	Purpose
GET	/	API health/status response
POST	/api/generate-pdf	Render document content to PDF
POST	/api/generate-html	Return generated HTML
POST	/api/generate-docx	Return HTML with a legacy Word MIME type

The generate-docx endpoint currently returns HTML content labelled as application/msword; it is not a full DOCX package generator. Do not describe it as native DOCX export.

Security and privacy

Document processing can involve sensitive personal, business or legal information. Before exposing the service publicly:

• Validate file type using content inspection, not only filename extensions or browser MIME values.
• Enforce upload size, page count, processing time and resource limits.
• Sanitize HTML and validate URLs and embedded resources before rendering.
• Restrict CORS to trusted origins.
• Avoid logging document content or sensitive metadata.
• Use temporary storage with cleanup if file-based processing is introduced.
• Isolate conversion workers and restrict network access where possible.
• Define retention and deletion behavior for uploaded and generated files.
• Never expose secrets in frontend environment variables or client bundles.

See Security.

Documentation

• Product definition
• Architecture
• Roadmap
• Architecture decisions
• Security
• Contributing
• Claude Code instructions
• Project implementation prompt

License

Released under the MIT License.

---

Part of the WI Arch ecosystem.