Agent Skills, Explained: An Open Standard Meets Graph Engineering in Memgraph

Agents are only as good as the skills and tools we give them. In this community call, we'll dive into Memgraph Skills for agents and humans, a practical approach to packaging graph expertise as reusable, composable skills. Initially developed by Anthropic, but currently available as an open standard: https://agentskills.io/home

We'll show how the skills can power autonomous agents and guide engineers in accomplishing tasks with Memgraph, such as writing a query module in Rust, Python, and C++, or even developing a GraphRAG pipeline.

In this session, you will learn:

• THow “skills” turn one-off prompts into repeatable, testable capabilities
• How to represent graph operations as skills an agent can use safely (and a team can maintain)
• How the same skills can guide engineers to ship faster across stacks (e.g., Rust, Python, C++ query modules)
• How skills plug into GraphRAG workflows to make multi-hop retrieval and reasoning more deterministic