Rankenstein
Rankenstein is an autonomous, self-correcting content engine for Shopify stores and agencies.
Core trick
Every article claim must match a FactsTable built from the store catalog, and an independent verifier refuses any ungrounded claim before review.
Description
Rankenstein is an autonomous, self-correcting content engine for Shopify stores and agencies. It takes a store from connect → grounded content → human review → live publish, fully autonomously between human checkpoints. The problem it solves: AI content tools hallucinate facts and publish unreviewed slop that erodes SEO and trust. Rankenstein never invents a fact — every claim is grounded in a FactsTable built from the store's real catalog, and an independent fresh-context verifier grades each piece against a rubric, refusing-and-flagging ungrounded claims (fabricated GSM, certifications, reviews) before a human ever sees them. Keyword research uses SERP-ownership scoring to pick fights a store can actually win, and output is structured for answer engines (AEO + JSON-LD), not just Google. Reviewers leave anchored comments that drive a surgical edit — only the commented span changes, proven by an independent span diff — then approve and publish to the live storefront with a version snapshot and one-click rollback. We connected the real EZ Fabric store, synced 903 products, and ran the full pipeline end to end on rankenstein.app
README excerpt
# Rankenstein **An autonomous, self-correcting content engine that never publishes without proof.** Rankenstein researches keywords it can actually win (with SERP-ownership evidence), writes articles and product content grounded *only* in verifiable source data, structures everything for search engines and answer engines (AEO), grades its own output against a machine-checkable rubric with independent verifier agents, and ships only after a human approves — with version snapshots and one-click rollback on every publish. Part of the [Jakka AI](https://www.jakka.ai) platform. Built for site owners and agencies. ## Status 🚧 **This repository is the Claude Fable 5 Build Day project (June 13, 2026).** All product code is being built during the event, in-session, by Claude Fable 5 running against the brief and rubric below. - [`inputs/BRIEF.md`](inputs/BRIEF.md) — the build brief given to the model - [`inputs/RUBRIC.md`](inputs/RUBRIC.md) — the machine-checkable definition of done (the model's verifier agents grade against this) - [`inputs/LAYER-CONTRACTS.md`](inputs/LAYER-CONTRACTS.md) — typed contracts for every pipeline stage - [`inputs/`](inputs/) — see its README for provenance (everything there predates the event) ## The loop (what makes this different) 1. **Ask first** — brand guidelines are drafted from a site crawl but a human confirms them; nothing generates from assumptions. 2. **Research first** — keyword choice requires SERP-ownership evidence (who owns the top