Documents

Add documents to your Context Repo workspace by typing content, uploading files in 75+ formats, or scraping web pages — then listen, search, version, and chat with them.

Documents are your knowledge base in Context Repo. You can add content three ways — type it directly, upload a file, or scrape a web page — and then edit, version, listen to, and discover related documents from your workspace.

Each document gets its own detail page with five tabs: Preview, Editor, Versioning, Similar Docs, and AI Chat.

Creating Documents

The documents toolbar offers three ways to add content to your workspace.

Manual Entry

Click Paste Content in the toolbar to open the creation dialog. This is the fastest way to add text or markdown content directly.

Field	Required	Description
Title	Yes	A descriptive name for your document
Content	Yes	The document text — plain text or markdown are both supported
Tags	No	Keywords to help organize and find this document later
Collections	No	Assign to one or more collections right away
Auto-classify	No	When enabled, Context Repo automatically adds the document to matching collections based on its content

Click Create Document to save. You'll be taken directly to the document detail page.

File Upload

Click Upload Files to open the upload dialog. You can drag and drop files, click to browse, or even capture a photo with your device's camera.

Context Repo supports 75+ file formats across six categories:

Category	Popular Formats
Documents	PDF, DOCX, DOC, RTF, Pages
Presentations	PPTX, PPT, Keynote, ODP
Spreadsheets	XLSX, XLS, CSV, TSV, Numbers
Images	JPG, PNG, GIF, SVG, WebP (with OCR text extraction)
E-Books	EPUB
Web Content	HTML, HTM

You can upload multiple files at once. Each file gets its own title, tags, and collection assignments — or you can apply the same collections to all files with the Apply to all toggle.

Processing Options

Expand the Processing Options section for fine-grained control over how your files are processed:

Document Language — Select the primary language (15 languages supported including English, Spanish, French, Chinese, Japanese, and more)
Parsing Mode — Choose the quality-vs-speed tradeoff:
- Default (Auto-detect) — Balanced processing for most documents
- Fast Mode — Quicker processing, lower credit cost
- Premium Mode — Highest quality extraction, uses more credits
- LLM Mode — AI-enhanced parsing for complex documents
Document Type Preset — Optimize parsing for specific types like emails, resumes, financial reports, or research papers
OCR Settings — Control text extraction from images and scanned PDFs:
- Auto-detect (recommended) — Enables OCR only when needed
- Always enabled — Extracts text from all documents
- Disabled — Skips OCR for faster processing

Click View all supported formats in the upload dialog for a complete list of every file type, organized by category.

Web Scraping

Click Scrape Page to extract content from any web page. Enter a URL and Context Repo uses Firecrawl to pull the page content into a document.

Field	Required	Description
URL	Yes	The web page address to scrape — `https://` is added automatically if you don't include it
Collections	No	Assign the scraped content to collections
Auto-classify	No	Automatically file into matching collections

Advanced Scraping Options

Expand the Advanced Options section for more control over the Firecrawl extraction:

Option	Default	Description
onlyMainContent	On	Strips navigation, headers, footers, and sidebars so you get just the article or page body
includeScreenshot	Off	Captures a visual snapshot of the page alongside the text content
waitFor	0 seconds	Set a delay (0–30 seconds) before scraping to let dynamic content load — useful for pages that render with JavaScript
extractStructured	Off	Pull specific data points from the page using a natural language prompt. For example, enter "Extract the product name, price, and description" to get structured fields from a product page

Editing and Preview

Open any document to see its detail page. The Preview tab shows your content rendered as formatted markdown, along with metadata like source URL, author, and usage stats (views, shares, exports).

Switch to the Editor tab to make changes:

Click the Edit button in the top-right corner
Update the title, content, or collection assignments
Click Save Changes when you're done

The editor supports full markdown with a live editing experience. If you change your mind, click Cancel to discard your edits.

For uploaded files, the Editor tab also displays processing information — the language detected, parsing mode used, OCR status, pages processed, and credits consumed (see Credit Calculation below).

Version History

Every time you save changes to a document, Context Repo automatically creates a new version. This works the same way as prompt versioning — no extra steps required.

What's Tracked

Each version records:

The full document content at that point in time
A change log describing what was updated
A timestamp
A line-by-line diff compared to the previous version, shown in both content and metadata tabs

Restoring a Previous Version

Switch to the Versioning tab to see the full timeline. Click the expand arrow on any version to view its diff, then click Restore to roll back. Restoring creates a new version with the old content — your history is never lost.

Text-to-Speech

The Preview tab includes a built-in audio player that converts your document to speech. Click Listen to Document to generate audio.

Text-to-Speech requires a Pro Max or Ultimate subscription. View plans for details.

Voice Options

Choose from six voices, each with a distinct character:

Voice	Description
Alloy	Neutral and balanced
Echo	Warm and conversational
Fable	Expressive British accent
Nova	Friendly and energetic (default)
Onyx	Deep and authoritative
Shimmer	Soft and gentle

Quality Tiers

Pick between two quality levels from the settings menu:

Standard — Faster generation, good for quick listening
HD — Higher fidelity audio for when quality matters

Playback Controls

The audio player supports:

Play, pause, skip forward/back 10 seconds
Adjustable playback speed (0.5x to 2x)
Volume control
Download the generated audio as an MP3
For longer documents, automatic chunking with part-by-part navigation and optional auto-advance

Credit Calculation

When you upload a file, Context Repo displays a credit breakdown on the document's Editor tab. This shows how many processing credits were used and the estimated cost based on the parsing mode:

Parsing Mode	Credits per Page
Fast	1 credit
Balanced (default)	3 credits
Premium	45 credits
LLM-Enhanced	45 credits

OCR adds 3 credits per page when enabled. If a document was already processed and cached, re-processing costs zero credits.

Credit calculation only applies to uploaded files. Manual text entries and scraped web pages don't use processing credits.

Managing Documents

The actions menu (three-dot icon) on any document detail page gives you quick access to:

Copy content — Copies the title and content to your clipboard
Export — Download as Markdown or JSON
Share — Copy a shareable link
Archive / Unarchive — Hide from your main list without deleting
Delete — Permanently remove the document and all its versions

Deleting a document removes it and all its version history permanently. If you might need it later, consider archiving instead.

AI Chat

Each document has a built-in AI chat interface in the AI Chat tab. Start a new session to ask questions about the document content, get summaries, or have the AI analyze specific sections. Your chat sessions are saved and you can return to them anytime.