bib-cite-web

name	bib-cite-web
description	Create bibliography citations from web page URLs with automatic Wayback Machine archival and metadata extraction. Use when the user asks to cite a website, create a citation for a URL, archive and cite a web page, or generate a bibliography entry from a web address.

Web Page Citation Creator

Create bibliography citations from web page URLs with automatic archival snapshot and metadata extraction.

Features

• Wayback Machine Integration: Automatically submits URLs to the Internet Archive for preservation
• Metadata Extraction: Extracts title, author, description, site name, and publish date from semantic HTML
• Multiple Formats: Outputs citations in BibTeX or CSL JSON format
• Smart Citation Keys: Generates citation keys from domain + author + year

Usage

npx tsx plugins/bib/scripts/cite-web.ts <url>
npx tsx plugins/bib/scripts/cite-web.ts <url> --format=bibtex
npx tsx plugins/bib/scripts/cite-web.ts <url> --no-wayback
npx tsx plugins/bib/scripts/cite-web.ts <url> --output=citations.bib

Metadata Extraction

The script extracts metadata from semantic HTML tags:

Title

• <title> tag
• Open Graph: <meta property="og:title">
• Twitter Card: <meta name="twitter:title">
• Standard: <meta name="title">

Author

• <meta name="author">
• Open Graph: <meta property="og:author"> or <meta property="article:author">
• Twitter Card: <meta name="twitter:creator">

Description

• <meta name="description">
• Open Graph: <meta property="og:description">
• Twitter Card: <meta name="twitter:description">

Site Name

• Open Graph: <meta property="og:site_name">
• <meta name="application-name">

Published Date

• Open Graph: <meta property="article:published_time">
• <meta name="publish-date"> or <meta name="date">

Arguments

• Positional argument: URL to cite
• --file <path>: Read URL from file (uses first line)
• --format <format>: Output format (default: bibtex)
  • bibtex or bib: BibTeX format
  • csl, json, or csl-json: CSL JSON format
• --no-wayback: Skip Wayback Machine submission (faster, but no archive)
• --output <file>: Write output to file (default: stdout)

Output Formats

BibTeX

@online{smithexample2024,
  author = {John Smith},
  title = {Example Article Title},
  url = {https://example.com/article},
  urldate = {2024-03-15},
  year = {2024}
}

CSL JSON

[
  {
    "id": "smithexample2024",
    "type": "webpage",
    "title": "Example Article Title",
    "author": [{"literal": "John Smith"}],
    "URL": "https://example.com/article",
    "accessed": {"date-parts": [[2024, 3, 15]]},
    "archive-url": "https://web.archive.org/web/20240315123456/https://example.com/article"
  }
]

Examples

Basic citation

npx tsx plugins/bib/scripts/cite-web.ts "https://example.com/article"

Output:

@online{example2024,
  title = {Example Article Title},
  url = {https://example.com/article},
  urldate = {2024-03-15}
}

With Wayback archival

npx tsx plugins/bib/scripts/cite-web.ts "https://blog.example.com/post"

Output includes archive URL:

@online{example2024,
  title = {Blog Post Title},
  url = {https://blog.example.com/post},
  urldate = {2024-03-15},
  note = {Archived at https://web.archive.org/web/20240315123456/...}
}

CSL JSON format

npx tsx plugins/bib/scripts/cite-web.ts "https://docs.example.com" --format=csl

Skip archival (faster)

npx tsx plugins/bib/scripts/cite-web.ts "https://example.com" --no-wayback

Save to file

npx tsx plugins/bib/scripts/cite-web.ts "https://example.com" --output=citations.bib

Batch processing

# Create file with URLs (one per line)
echo "https://example.com/article1" > urls.txt

# Cite each URL
while read url; do
  npx tsx plugins/bib/scripts/cite-web.ts "$url" >> citations.bib
done < urls.txt

Citation Key Generation

Citation keys are automatically generated from:

1. Domain name: example.com → example
2. Author (if available): John Smith → smith
3. Year: Archive date or publish date or current year

Examples:

• https://blog.example.com/post by John Smith (2024) → smithexample2024
• https://example.com/article (no author, 2023) → example2023

Wayback Machine Integration

By default, the script submits URLs to the Internet Archive's Wayback Machine for preservation:

1. Submission: Sends URL to https://web.archive.org/save/<url>
2. Archive URL: Extracts the permanent archive URL from response
3. Archive Date: Records the snapshot timestamp
4. Fallback: If submission fails, continues without archive

The archive URL is included in the citation:

• BibTeX: In note field or custom archiveurl/archivedate fields
• CSL JSON: In archive-url field

Skip archival with --no-wayback for faster execution when archiving isn't needed.

Error Handling

The script handles various error scenarios:

• Invalid URL: Validates URL format before processing
• Fetch failures: Reports HTTP errors with status codes
• Missing metadata: Falls back to "Untitled" for missing titles
• Wayback failures: Continues without archive if submission fails
• No author: Omits author field if not found

Errors are written to stderr, while citations are written to stdout (or file).

Limitations

• JavaScript-heavy sites: May not extract metadata from dynamically rendered content
• Paywalls: Cannot access content behind authentication
• Rate limiting: Wayback Machine may rate-limit submissions
• No PDF support: Only HTML pages (use separate tool for PDFs)
• Simple parsing: Uses regex matching, not full DOM parsing

For complex pages or JavaScript-rendered content, consider:

1. Using --no-wayback to skip archival
2. Manually editing the citation after generation
3. Using browser developer tools to inspect metadata tags

Related Skills

• bib-create: Create bibliography entries interactively
• bib-read: View existing bibliography entries
• bib-convert: Convert between bibliography formats
• wayback-submit: Submit URLs to Wayback Machine without citation generation