---
title: "Pattern Catalog"
description: "Eight patterns that keep AI product structure fast to build and resistant to collapse."
canonical: "https://architecture.reopt.ai/patterns"
language: "en"
last-updated: "2026-10-20"
alternate-markdown: "https://architecture.reopt.ai/ko/patterns.md"
site-index: "https://architecture.reopt.ai/llms.txt"
source: "reopt architecture"
---

# Pattern Catalog

> Eight patterns that keep AI product structure fast to build and resistant to collapse.

Each pattern answers a concrete judgment question a team faces. Responsibility and ownership, collaboration and coordination, trust and verification — three areas where structure tends to break.

## What problem are you solving?

Find the question below that matches your current concern and jump to the pattern.

### Responsibility & Ownership

#### Responsibility Partitioning

Name the owner of each outcome and draw the boundary around it.

- **Judgment question:** Is this responsibility independent enough to live in its own agent?
- **OCLS:** OWN
- [responsibility-partitioning](https://architecture.reopt.ai/patterns/responsibility-partitioning.md)

#### Module Contract

Declare the conditions, authority, and failure paths of every execution unit.

- **Judgment question:** Under what conditions must this module refuse or fail?
- **OCLS:** CONTRACT
- [module-contract](https://architecture.reopt.ai/patterns/module-contract.md)

### Collaboration & Coordination

#### Context Routing

Design information flow so each participant receives exactly what it needs.

- **Judgment question:** What must this agent not be unaware of right now?
- **OCLS:** LAYER
- [context-routing](https://architecture.reopt.ai/patterns/context-routing.md)

#### State and Memory Control

Separate short-term state from long-term memory to prevent leakage and confusion.

- **Judgment question:** What stays in the session, and what gets promoted to system memory?
- **OCLS:** LAYER
- [state-and-memory-control](https://architecture.reopt.ai/patterns/state-and-memory-control.md)

#### Cost Control

Manage token budget, model selection, and call frequency structurally to control the cost curve.

- **Judgment question:** Does this reasoning step really need this model?
- **OCLS:** LAYER
- [cost-control](https://architecture.reopt.ai/patterns/cost-control.md)

### Trust & Verification

#### Evaluation and Guardrails

Distinguish acceptable judgment from dangerous judgment via evaluation criteria and safety rules.

- **Judgment question:** By what criteria do we measure quality and risk?
- **OCLS:** SHARPEN
- [evaluation-and-guardrails](https://architecture.reopt.ai/patterns/evaluation-and-guardrails.md)

#### Human Approval

Keep high-cost, high-risk, high-impact decisions inside a human-approval flow.

- **Judgment question:** When should the agent stop and hand off to a human?
- **OCLS:** SHARPEN
- [human-approval](https://architecture.reopt.ai/patterns/human-approval.md)

#### Decision Traceability

Record judgment rationale, choices, and collaboration paths as structured logs.

- **Judgment question:** Can the rationale of this agent's decision be explained after the fact?
- **OCLS:** SHARPEN
- [decision-traceability](https://architecture.reopt.ai/patterns/decision-traceability.md)

## Pattern Relationship Map

The eight patterns are not independent. Responsibility and ownership determine the collaboration structure, and the collaboration structure produces the demands of trust and verification. Patterns connected in the map below have an outsized effect when applied together.

---

_Markdown representation of https://architecture.reopt.ai/patterns. Full index: https://architecture.reopt.ai/llms.txt HTML original: https://architecture.reopt.ai/patterns · 한국어: https://architecture.reopt.ai/ko/patterns.md_