Skip to content

THE LEARNING LIBRARY

AI, with the pieces
put together.

Understand the vocabulary. Explore the architecture. See how the concepts show up in working systems.

YOUR STARTING POINT

New to this? Start with the connections.

Five foundations, in a useful order. No background in engineering required.

Start with APIs & connectors

01 / FOUNDATIONS

Get comfortable with the components.

01 / 09 APIs, MCPs & Connectors API A DEFINED REQUEST AND RESPONSE APPLICATION THE AI APPLICATION CONNECTOR AUTH · FIELDS · RETRIES PERMISSIONS DECIDE WHAT PASSES CRM SYSTEM OF RECORD MCP · ONE WAY TO DESCRIBE WHAT IS THERE THE REQUEST, AND THE DATA IT RETURNS /LEARN/APIS-MCPS-CONNECTORS 01

APIs, MCPs & Connectors

How AI reaches your company’s tools and information.

Read the guide → 02 / 09 Tools & Skills TOOL CALL READ WRITE A CAPABILITY. IT CAN BE INVOKED. SKILL GOLD STANDARD 01020304 A PROCEDURE. IT DECIDES ORDER. SAME TOOL. DIFFERENT PROCEDURE. THE PROCEDURE CALLING THE CAPABILITY /LEARN/TOOLS-AND-SKILLS 02

Tools & Skills

The difference between a capability and a procedure.

Read the guide → 03 / 09 AI Agents THE JOB CONTEXT MODEL TOOL 01TOOL 02TOOL 03 BOUNDARY · ALLOWED ACTIONS ONLY STOP CONDITION OR HAND TO A PERSON OBSERVE · CHOOSE · ACT THE JOB MOVING THROUGH THE AGENT /LEARN/AI-AGENTS 03

AI Agents

Give an agent a job, the right context, and a boundary.

Read the guide → 04 / 09 Workflows & Automations TRIGGER ADDED LAST AFTER THE STEPS CONNECT 01020304 INTAKEDRAFTCHECKSEND EACH STEP HANDS THE NEXT A USABLE INPUT NAMED PERSON THE FAILED PATH HAS AN OWNER THE WORK MOVING BETWEEN STEPS /LEARN/WORKFLOWS-AUTOMATIONS 04

Workflows & Automations

Connect the steps before you hand them a trigger.

Read the guide → 05 / 09 Context & Memory CONTEXT WINDOW WHAT THE MODEL SEES, THIS TURN FALLS OFF THE EDGE MEMORY WHAT THE SYSTEM KEEPS APPROVEDFACTSDECISIONSQUESTIONS RETRIEVED APPROVED WORK IS WRITTEN BACK WHAT MOVES IN AND OUT OF CONTEXT /LEARN/CONTEXT-AND-MEMORY 05

Context & Memory

What the model can see, and what the system remembers.

Read the guide →

02 / ARCHITECTURE

See the system behind the tools.

06 / 09 The Component Map ONE OPERATING SYSTEM THE COMPANY RUNS ON EMPLOYEE ADOPTIONBUSINESS USE CASESCOMPANY KNOWLEDGETECHNOLOGY THE VALUE REALIZEDTHE PURPOSETHE CONTEXTTHE FOUNDATION THE CONNECTION IS THE WORK OVERINVEST IN ONE AND IT STALLS THE VALUE, WHEN ALL FOUR CONNECT /LEARN/COMPONENT-MAP 01

The Component Map

See how the layers of a company AI system fit together.

Read the guide → 07 / 09 The Operating Loop 01020304050607 ASK ENHANCE ROUTE GROUND BUILD SCAN APPROVE PASS OR FAIL, NOT AN OPINION ONLY A PERSON MOVES WORK HERE THE BRAIN APPROVED WORK RETURNS ONE REQUEST, ALL SEVEN STEPS /LEARN/OPERATING-LOOP 02

The Operating Loop

From an incoming request to a reviewed, reusable result.

Read the guide → 08 / 09 Graph Engineering ONE LARGE JOB TOO BIG TO CHECK AT ONCE NODENODENODENODENODE RESULT ITS OWN CHECK EACH NODE OWNS ONE OUTPUT THE JOB, ONE STEP AT A TIME /LEARN/GRAPH-ENGINEERING 03

Graph Engineering

Split a large job into connected, accountable pieces.

Read the guide → 09 / 09 Guardrails & Gates WRITTEN RULE ONE SOURCE PER FACTNO BANNED TERMSUNKNOWN IS FILED THE RULE BECOMES THE CHECK THE GATE PASS OR FAIL, NOT AN OPINION CHECK DRAFT BACK TO THE BUILD STEP THE PATH IT DID NOT TAKE APPROVED ONLY A PERSON MOVES WORK HERE FAILED WORK GOES BACK, NOT FORWARD THE DRAFT, STOPPED AT THE CHECK /LEARN/GUARDRAILS 04

Guardrails & Gates

Turn a written rule into a check you can prove works.

Read the guide →

03 / PUT IT TO WORK

Understanding AI is the starting point.

AI in Practice covers the Company OS, the Method, the Intake, and adoption.

Explore AI in Practice