Knowledge Base Schema
The shared brain every agent reads from.
What it stores
The shared brain every agent reads from. It holds the knowledge entries, the rules that decide what to load for a given task, and the authoritative agent registry — the same registry this portfolio is built from.
Owned by the Chief of Staff. Lives in the Internal Tools database · brain schema.
The tables
Friendly names first, with the real table name underneath. Key fields only.
Knowledge Entriesbrain.entries
One entry per concept, glossary term, or standing rule — the canonical reference agents read.
| Column | Type |
|---|---|
slug | Text |
title | Text |
summary | Text |
content | Text |
versionBumped on every edit | Number |
Relationships
- one-to-one
brain.categories. Filed in a category - many-to-many
brain.tags. Tagged for discovery
Load Rulesbrain.load_rules
Maps a task keyword to exactly the knowledge an agent should load before starting it.
| Column | Type |
|---|---|
task_keyword | Text |
required_slugsMust-load entries | List |
optional_slugsNice-to-have entries | List |
Agent Registrybrain.agents
The authoritative directory of every specialist agent — the source this portfolio is generated from.
| Column | Type |
|---|---|
name | Text |
workspace | Text |
summary | Text |
scope | Text |
mcp_serversConnectors the agent loads | List |
Categoriesbrain.categories
The fixed set of knowledge categories.
| Column | Type |
|---|---|
label | Text |
description | Text |
Relationships
- one-to-many
brain.entries. Groups many entries
Tagsbrain.tags
Cross-cutting tags for finding related knowledge.
| Column | Type |
|---|---|
label | Text |
domain | Text |
Relationships
- many-to-many
brain.entries. Applied across entries
Entry Linksbrain.entry_relations
A small graph of how knowledge entries relate to one another.
| Column | Type |
|---|---|
relation_typeHow the two relate | Text |
Relationships
- many-to-many
brain.entries. Connects entry to entry
You don't need all of this to start
Most of what these agents rely on has a simpler equivalent. Anywhere we use a database, you can usually start with a spreadsheet or a document — whatever you're comfortable with. Anywhere we built a custom MCP server, you can reach for a prebuilt AI connector instead of building your own.
We go further because it's where our experience pays off. Years of hands-on database work make a custom data layer feasible for us, and purpose-built MCP servers let us hand each agent a short, sharp instruction file and exactly the tools it needs — custom descriptions, only the endpoints we want, and the occasional extra gate — instead of a long manifesto or playbook. Nothing more, nothing less.
Ready to build technology that works for your mission?
Tell us where your organisation is and what's slowing your team down. We respond personally.