Tool Gate · How it works
How the Data-Model Sanity Bench gate works
How the Data-Model Sanity Bench rolls ten tables into one verdict: a prerequisite gate drops any table holding records with no owner before uniqueness or types are even read - the sample reads NOTHING TO BUILD ON.
How the gate works
This tool grades each table on whether the database enforces its rules or you are quietly enforcing them by hand. It reads 10 tables and rolls them into one verdict — the worst, never the average — and then a gate that can only make it worse.
The 10 tables
| Table | Verdict |
|---|---|
| T-01 · User profiles | DATABASE ENFORCES ITreal reference · unique enforced · proper types |
| T-02 · Orders and their totals | YOU ENFORCE ITmoney stored as text or number |
| T-03 · Uploaded documents | NOTHING TO ENFORCE AGAINSTholds records with no owner at all |
| T-04 · Comments on documents | YOU ENFORCE ITowner is a plain value, not a reference |
| T-05 · The plan tiers you offer | NO RULES NEEDEDa reference list, not records |
| T-06 · Subscriptions | YOU ENFORCE ITuniqueness assumed, never enforced |
| T-07 · Country and currency codes | NO RULES NEEDEDa reference list, not records |
| T-08 · File upload recordsfix first | NOTHING TO ENFORCE AGAINSTowner not found |
| T-09 · Team invites | YOU ENFORCE ITuniqueness answer could not be found |
| T-10 · Activity log | DATABASE ENFORCES ITreal reference · not assumed · proper types |
The verdict bands
A table that holds records with no owner drops straight to NOTHING TO ENFORCE AGAINST, before uniqueness or types are even read. There is no point grading the rules on a table that has nothing to attach a rule to — the ownership question is the prerequisite, and everything downstream is moot until it is answered.
In the worked sample it reads NOTHING TO BUILD ON — Two of ten tables hold records with nothing to enforce against, so the model has no floor. Four more work only because your code remembers to: money kept as a loose number, an owner stored as a plain value, uniqueness assumed rather than enforced. Two reference lists genuinely need no rules and never count against you. The one thing to fix first: T-08, file upload records — the owner answer could not be found, so it resolves to none and the prerequisite gate fires.
Common questions
Can every table look survivable on the How the Data-Model Sanity Bench gate works and the company still read NOTHING TO BUILD ON?
Yes. The roll-up takes the worst of 10 tables — never the average — then a worsen-only gate, the Prerequisite gate, escalates when trouble clusters. In the worked sample it reads NOTHING TO BUILD ON because two of ten tables hold records with nothing to enforce against, so the model has no floor. Four more work only because your code remembers to: money kept as a loose number, an owner stored as a plain value, uniqueness assumed rather than enforced. Two reference lists genuinely need no rules and never count against you.
What is a worsen-only gate?
A dispositive rule that can only lower a verdict, never raise it. One fatal flaw overrides an otherwise-set of survivable lines, because a single disqualifying gap shouldn't hide behind an average of the healthy ones. The tool also names the one thing to fix first — here, "T-08, file upload records — the owner answer could not be found, so it resolves to none and the prerequisite gate fires".
Grades how a schema is shaped, never a person. It connects to no database — it reads back what you tell it, using rules you can see.
Embed this diagram
Free to share and embed with attribution (CC BY 4.0) — keep the link to redhub.ai.
Interactive — renders the live diagram
<iframe src="https://redhub.ai/visuals/tool/data-model-sanity-bench.html" title="How the Data-Model Sanity Bench gate works — RedHub AI" width="760" height="2214" loading="lazy" style="border:0;width:100%;max-width:760px"></iframe>
<p style="font:14px/1.5 system-ui,sans-serif"><a href="https://redhub.ai/visuals/tool/data-model-sanity-bench">How the Data-Model Sanity Bench gate works</a> — by <a href="https://redhub.ai">RedHub AI</a>, the AI that tells you when to doubt it.</p>Image + link — a static picture for any blog
<a href="https://redhub.ai/visuals/tool/data-model-sanity-bench"><img src="https://redhub.ai/visuals/tool/data-model-sanity-bench-share.png" alt="Diagram of the Data-Model Sanity Bench: ten tables, a prerequisite ownership gate, and a sample reading NOTHING TO BUILD ON." width="760" loading="lazy" style="max-width:100%;height:auto;border-radius:16px"></a>
<p style="font:14px/1.5 system-ui,sans-serif"><a href="https://redhub.ai">RedHub AI</a> — the AI that tells you when to doubt it.</p>
This is how Data-Model Sanity Bench works. More diagrams in the Visual Field Guide.