Advanced · Systems

Graph Engineering

Visual diagram · Math · (in preparation) · Worked example · 3 difficulty levels.

TL;DR. Organizing multiple agents, work structures, and runtime states as one coordinated system — the layer above the individual agent harness.

Technical Definition

Organizing multiple agents, work structures, and runtime states as one coordinated system — the layer above the individual agent harness.

How it works

Introduced by the 2026 survey 'Graph Engineering in the Era of LLM Agents' (arXiv:2608.21156), graph engineering names the discipline that begins where harness engineering ends. Prompting and context shape a single model call; a harness and loop turn that model into one persistent agent; graph engineering decides how many units of execution, responsibility, and state should be organized as one system — which agents exist, how they hand off work, where shared state lives, and how the whole structure is observed and governed. It reframes multi-agent design from 'add more agents' to 'design the graph': nodes are agents or work structures, edges are responsibility and data flow.

Visual Explanation (flowchart)

Agent node ↔ shared state node ↔ agent node → orchestrated as one system graph with defined edges

Related Concepts

• Agent Harness — Everything surrounding a model that makes it useful in real workflows — the tools, permissions, memory, context, and orchestration around the model itself.
• Procedural Graphs — Self-evolving execution structures that organize what-to-do knowledge into (procedure, relation, procedure) triplets, guiding an agent's next action at each step.
• Agent orchestration — Agent orchestration is the management of multiple AI agents or LLM calls to efficiently handle complex tasks.
• Multi-Agent Systems (MAS) — Systems composed of multiple interacting intelligent agents that cooperate or compete to achieve individual or collective goals.