Wayback Machine History Extractor

Extract historical snapshots of any URL from the Internet Archive's Wayback Machine via the public CDX API. Maps to "what did this page look like in [year]?", "track competitor's homepage messaging over time", "audit my brand's historical claims" briefs.

Built 2026-05-03 as Demo #24.

Run

. ~/freelance/.venv/bin/activate
cd ~/freelance/portfolio_demos/wayback_history_extractor
python extract.py --target https://www.anthropic.com --from 20230101 --to 20260501 --limit 8

Result

• 8 unique-digest snapshots from anthropic.com (Jan-Feb 2023) ✅
• CDX collapse=digest server-side de-dupes identical content snapshots ✅
• Per-snapshot fields: date, snapshot URL, title, h1, meta-description ✅
• 1.5s politeness sleep — Wayback is a free public service ✅
• Tenacity retries with exponential backoff (4-20s) ✅

Why this brief class matters

Three real Upwork-style use cases:

1. Competitor messaging archaeology — "What did Stripe's homepage say in 2018 vs 2024?"
2. Compliance / fact-check — "Did this site claim X before regulators flagged it?"
3. SEO / content history audit — "What did our title tag look like in 2022 when we ranked #1?"

The CDX API is free, has no quota, and returns up to millions of snapshots per URL.