§ Executor · The structured-data layer of your site
SchemaForge.
Generates and maintains structured data across every page on your site. Lifts AI-search and rich-result visibility without engineering cycles, plugin drift, or quarterly audits.
01 What it does
SchemaForge is the platform that gives every page on your site a complete, accurate, machine-readable identity. It generates the structured data on publish, keeps it synchronized as content changes, and audits the whole site continuously for drift.
The output looks the same to a human reader and entirely different to a search engine, an LLM crawler, or an AI assistant trying to summarize your site. The latter three are now the entire growth surface. SchemaForge is what makes you legible to them.
Schema generation on publish
Every new page gets the right schema variant for its content shape, picked from the full schema.org catalog and generated against the body itself — not a one-size-fits-all template.
Continuous re-sync
When content changes — title edits, body rewrites, image swaps, author changes — schema gets re-emitted to match. No stale data pointing at last year's headline.
Entity cross-poll
Schema pulls validated entity metadata from your hub system (HubForge) so people, places, orgs and things resolve to canonical IDs across the whole site.
Site-wide audit loop
A standing crawl that detects missing, malformed, or outdated schema across the corpus — and ships fixes through the same pipeline that publishes them.
02 How it works
SchemaForge runs as durable workflows in production cloud. Two triggers — publish events and a periodic audit — drive the entire surface.
Listen. The platform subscribes to publish and update events from your content system. Every new asset or edit lands as an event, with the body, metadata and entity tags attached.
Classify. A small classifier picks the right schema variant for the content shape — whatever your domain demands, from product pages to case studies to event listings — and routes to the appropriate builder.
Build. The builder emits a complete structured-data object with every field the variant needs — headline, source, dates, media, entity references, body markup, ordering — picked per content shape, not stamped from a template.
Write & verify. Schema gets written into the page head via your CMS's structured-data hook. A verification pass reads it back to confirm it's live and well-formed.
Audit. A site-wide sweep crawls the corpus on cadence, finds drift (missing schema, broken refs, outdated entity IDs) and queues fixes. Every fix lands as a new write event — the loop is closed.
03 What it replaces
SchemaForge collapses the entire structured-data function — usually scattered across plugins, devs, and quarterly audits — into a single operated layer.
04 Live receipts
SchemaForge currently operates the structured-data layer of a live reference deployment — an editorial publishing platform under Umbra Group. Receipts below are verifiable in the audit log.
Reference deployment — sustained since 2026-05
- 100% schema coverage across 1,800+ posts, sustained for months. Every new publish gets schema; every edit re-syncs.
- Six+ schema variants in production, picked per content shape from the full schema.org catalog as the corpus demands.
- Loop closed end-to-end with OnPage Doctor: audit findings route into the schema pipeline automatically.
- Zero engineering tickets raised for schema work since deployment. Every new variant is added in the platform, not the codebase.
- Entity cross-poll live with HubForge — 240+ entities resolve to canonical hub IDs across every article that references them.
Source — Internal observatory + audit · 2026-05
05 Pairs with
SchemaForge is most powerful when paired with the platforms it shares data with.
HubForge
Provides canonical entity IDs that SchemaForge references in every article's structured data.
Pairs with — on-page auditOnPage Doctor
Audit findings about missing or malformed schema route directly into the SchemaForge fix pipeline.
Operated by — cockpitAtlas
Schema run failures and drift alerts surface in the Atlas brief and watcher channel.
Feeds — distributionPressRelay
Distribution channels can pull validated schema metadata when generating captions and previews.
06 Deploy in your org
SchemaForge is one of the executors deployed during a Lighthouse Sprint. The Sprint maps your content variants and CMS shape; SchemaForge is built against that map.
Discovery (Pulso). We crawl your site, classify the content variants you publish, and produce the schema matrix — which variants exist, which are missing, which are wrong.
Build (Sprint). Variant-specific builders are written against your content. The publish-event hook is wired into your CMS. Audit loop is configured at your preferred cadence.
Ship (Watch). Platform goes live behind Atlas. Coverage climbs to 100%. Drift is caught in cycles, not quarters. Your structured-data function is now operated, not staffed.
07 Pricing line
SchemaForge is included as an executor in any Lighthouse Sprint that touches your SEO and discovery workflow. Operating cost is bundled into your Lighthouse Watch tier.
Standalone deployment (SchemaForge alone, without other executors) is available from the base Sprint tier. Pricing is documented in the Sprint framework.
08 Engage
See 100% coverage on a real site, end to end.
We walk you through the live reference deployment: schema generation on a fresh publish, the audit loop catching a deliberately broken entry, and entity cross-poll resolving in real time.