---
title: "The knowledge graph · every agent, model and company connected · Agent Fieldbook"
url: https://agentfieldbook.org/kg/
description: "Every connected entity in the Fieldbook rendered as one 3D knowledge graph, clustered by category and sized by number of relationships. Drag to orbit, hover to name a node, click any node to open its sourced record."
section: "Overview · Knowledge Graph About"
source: Agent Fieldbook — generated from the published page
---

# The corpus as one connected graph

**The Fieldbook is a connected corpus, not a static directory.** This graph brings models, agents, orchestrations, frameworks and their relationships into one explorable system. It shows how the categories in the dataset converge, and where the most connected ideas sit.

## Field notes

Agents, models, companies, supply nodes, memory and topologies, in one space

**Relationships are as useful as records.** Each node represents part of the corpus, clustered by category. Node size reflects the number of relationships it holds, which makes the most connected concepts and systems easier to spot.

Zoom in, zoom out, hover and open a node to move from a high-level pattern to its underlying records and source material. The goal is not to display a collection of data points, it is to reveal the structure that emerges between them.

- [Brandon Chaplin](https://www.linkedin.com/in/brandon-chaplin-digital-marketing-strategist)

## Common questions

### What is a knowledge graph?

A way of storing information as things and the links between them. Each thing is a node, each link is an edge with a named type, such as 'built by' or 'runs on'. Because the links are stored explicitly, you can follow a path across the data instead of only looking rows up.

### What are nodes and edges?

Nodes are the things: a company, a model, a chip, an idea. Edges are the named relationships between them, and they usually point in one direction. Node size normally shows degree, meaning how many relationships that node holds.

### What is the difference between a knowledge graph and a database?

A relational database stores rows in tables and joins them when you ask. A knowledge graph stores the relationship itself, so connections several steps apart are cheap to follow. Use a database for known, repeated queries and a graph when the question is about how things connect.

### Why do AI agents use knowledge graphs?

They give a model facts it cannot invent. An agent can look up an entity, follow its relationships and cite where each fact came from, rather than relying on what the model remembers. They also make gaps obvious, because an unconnected node stands out.

### What is GraphRAG?

Retrieval-augmented generation that reads a knowledge graph rather than a pile of text chunks. Instead of fetching passages that look similar to your question, it walks the relationships around the relevant entities. It tends to do better on questions that need several facts joined together.

## Next in the learning path

- [Network The physical supply chain](https://agentfieldbook.org/network/)

- [Provenance The citation layer beneath](https://agentfieldbook.org/provenance/)

- [All agents The catalogue in full](https://agentfieldbook.org/catalog/all/)

- [Models The model landscape](https://agentfieldbook.org/software/models/)

- [Topologies How agents are shaped](https://agentfieldbook.org/design/topologies/)