Vector Entities skill

The skill file

---
name: vector-entities
description: Map the entities a page must contain to align semantically with a query — base terms, quantified proof, multimodal formats, and the expert concepts competitors omit.
owner: Moonrank GEO Agent
---

# Vector Entities

Modern retrieval compares meaning, not string matches. A page is chosen because
its vector sits close to the query's. This maps what has to be inside the page
for that to happen — and, just as importantly, what would make a model remember
it rather than merely match it.

## When to use
- Before writing a new page, as the input to the brief.
- On a page that ranks but is never cited or quoted.
- To understand the semantic gap against competitors.

## Inputs
- **Required:** the target query, its intent type, and the audience's expertise level.

What is the Vector Entities skill?

The Vector Entities skill is how Moonrank's GEO Agent works out what a page must contain to be selected by a system that compares meaning rather than matching strings. It produces four columns of ten: the technical entities that are the entry ticket, the quantified proof that raises a model's confidence, the formats the intent expects, and the expert concepts almost nobody includes.

The fourth column is the one that matters most and is hardest to fake. Its test is simple: if a competitor could copy the concept in five minutes, it isn't divergence. Real divergence comes from your own data, your own delivery and your own failures — which is exactly why it makes a model retain your page rather than average it into everyone else's.

It's written as a SKILL.md, a plain-language instruction format that people and AI agents can both follow. Read it as the input to a brief, hand it to a writer, or let the GEO Agent produce it for a target query.

See it in action