Agent Design Patterns Society

A public pattern language for production AI agent systems.

ADPS is an independent research community developing shared technical references for production AI agent systems. It brings together software design research, open-source harness engineering, and attributed enterprise field reports, then turns recurring problems into pattern specifications that engineers can compare, test, and cite.

The 28-pattern matrix — 7 cognitive functions × 6 execution topologies

The 28-core-pattern matrix · 7 cognitive functions × 6 execution topologies

What we do

ADPS publishes three connected sets of technical references:

The framework

The two-axis framework — cognitive function × execution topology — was introduced in A Two-Dimensional Framework for AI Agent Design Patterns (Huang & Zhou, arXiv:2605.13850) and developed further in the Manning book Designing AI Agents. ADPS now stewards the catalog as a community asset.

Browse the pattern white papers →

Founders

The founding team works across agent architecture, software engineering, open-source infrastructure, and enterprise systems.

Read the full team profiles →

Advisor

Expert Group

Read full profiles →

Assets

Reach out

ADPS is open to engineers and organizations with field experience building production-grade agent systems — at scale, in regulated industries, or under hard reliability constraints. Membership is by contribution, not application.

Fill the three fields below and we'll get in touch through WeChat. Nickname is fine — no real name required.

Prefer GitHub? Open a discussion or issue on agent-design-patterns.