What this project ships
- Publish the harvested backlog. The pipeline has crawled, classified, and extracted many franchises beyond the first one that is live end-to-end. They sit extracted-but-unpublished. This project runs the publish step — building the per-franchise read bundles the public site serves — so each franchise becomes a browsable surface rather than warehoused data.
- Make each franchise whole on arrival. A published franchise should land with its canonical pages, its graph connections (P-026), and the attribute QA already applied, so coverage grows in complete units, not half-populated ones.
Why this matters now
The encyclopedia's value scales with breadth: one franchise proves the model, fifty makes it a destination and a discoverability surface. The expensive work — harvesting and extraction — is already done for these franchises; what stands between them and the public is the publish step. Closing that gap is one of the highest-leverage moves available, because it converts work already paid for into live, indexable, monetizable pages.