Data Sources

Every source of truth the content engine reads from — where it lives, what it provides, and how often it refreshes.

← Contents ⬇ Download full PDF Glossary & Index

Every source the engine reads

Source	Provides	Location	Refresh
Network site list	Each site's domain + verified description → site features & spotlights	main:/root/wt-sites.json	As the network changes
Marketing-skills library	47 skills → AI-marketing tips	skills-data.js	Weekly
Quality-score snapshots	Daily AI-readiness scores → milestones & new-site spotlights	main:/root/ama-snapshots/	Daily
The built feed	The combined, cleaned, publishable items	hub:/opt/portal/listings/wholetech.json	Nightly
Connected-account tokens	Credentials to publish to each network	hub database	On connect/refresh

One principle governs all sources

Verified in, or held out.A field only becomes part of a post if it is present in the source. Truncated descriptions are dropped; missing scores are skipped; unverifiable facts are held. The feed is deliberately conservative — it would rather post less than post something it can't stand behind.

Adding a new source

To feed a new kind of post, add a section to gen-social-feed.py that reads the new data, cleans it, and emits items tagged with a new kind; then add matching templates. Keep the same rule: emit an item only when every field a template needs is verified.