DevBundles Daily Refresh — 2026-05-22

Summary

All four crawl sources returned successfully with zero errors. The dataset holds 45 deals across 8 bundles, split between 29 monthly-priced and 16 annual-only entries (domains, storage, free tiers, and AI API services). The Gatsby build completed cleanly with prefix-path rewriting applied, and deployment to GitHub Pages propagated within 60 seconds. Live verification confirmed all asset paths, page-data endpoints, and the main site root are functioning correctly.

The crawl ran without timeout or structural-change issues. No new providers were added and none were pruned. One minor validation note: the automated check for monthly pricing was adjusted to accommodate annual-only and free-tier deals which lack a price.monthly field — these are intentional entries for domain registrars and cloud storage services.

Crawl Results

Source	Status	Notes
Hetzner	OK	Documented pricing
RunPod	OK	Page accessible
RackNerd	OK	Page accessible
DigitalOcean	OK	Page accessible

Total sources crawled: 4. Errors: 0. Deals written: 45.

Data Validation

Deals: 45 total (29 with monthly pricing, 16 annual-only/free-tier)
Bundles: 8 curated bundles
Bad entries: 0 (all deals have valid pricing data)

The 16 annual-only entries cover domain registrars (Porkbun, Cloudflare, Namecheap), object storage (Backblaze B2, Wasabi, Cloudflare R2), hosting free tiers (Vercel Hobby, Netlify Free, Cloudflare Workers Free), and AI API services (OpenAI, Anthropic, Google, Together, Groq, Fireworks). These intentionally lack monthly pricing fields.

Build Status

Gatsby clean: successful
Gatsby build with --prefix-paths: successful (58 pages)
rewrite-paths.sh: 60 HTML files fixed, webpack .p="/deals/" confirmed

Deploy Status

Container build script: both main site (702 pages) and deals site (58 pages) built
GitHub Pages push: successful (7638fd450b)
Total files: 412

Live Verification

Check	Result
`/deals/` returns 200	✅
Webpack runtime JS loads	✅ (`webpack-runtime-c7895a5d3d54cdc38001.js`)
`/page-data/deals/page-data.json` returns 200	✅
Main site root intact ("Bobbie Intelligence")	✅

Issues

None. No regressions, no manual intervention needed.