Agent Memory in 90 Minutes
Build an ontologically sound agent memory graph system, give any agent the power to record memories and decision traces, use a memory-enabled agent to build a custom agent, and distil it all as a skill.
In this 2-hour workshop, you will learn
In this 90-minute workshop, approaching agent memory in three ways, you will:
- Build a fully functional, memory-backed agent with
create-context-graph. - Give Copilot access to
neo4j-agent-memoryover MCP, and build a more robust memory graph. - Use your MCP-backed Copilot to build a custom
neo4j-agent-memorycoding assistant.
At the end of this course, you will have the opportunity to distil your agent's memories from the graph into a persistent SKILL.md file with aip (Agent Instruction Protocol).
Throughout each exercise, you will come to understand:
- What
neo4j-agent-memoryis. - How to use
neo4j-agent-memoryin a way that suits your workflow. - How to make an agent to record memories according to a coherent agent-memory ontology.
GraphAcademy provisions your Neo4j instance and your AI access for the workshop, so there is nothing to sign up for beyond a GitHub account.
- A GitHub account, for the Codespace
- No Neo4j or Python experience is assumed
Agent Memory
Context Graphs
Ontologies
POLE+O
What you need to take part.
A GraphAcademy account
The workshop is delivered here on GraphAcademy, so you need to be signed in to work through the lessons and keep your progress. Creating an account is free.
A GitHub account, or Git on your own machine
You write code against the workshop-agent-memory-90 repository. The quickest route is a GitHub Codespace, an online editor that clones the code and installs everything for you — that needs a GitHub account.
If you would rather work locally, clone the repository with Git and run it in your own editor instead. No GitHub account is needed for that.