Skip to main content
Open Framework for Agentic Software Engineering

Your AI agent is only as good as the system around it.

The Agentic Engineering Framework is an open framework for building the agentic layer (prompt templates, quality gates, and workflow config) that powers an agentic development workflow, giving AI coding sessions the structure of a repeatable engineering process.

The Agentic Engineering Framework (AEF) is a framework, not a tool. It defines the patterns: prompt templates, quality gates, feedback loops, workflow config. You build the implementation that fits your stack. Browse the samples repository for prompt templates, alternative implementations of each layer, and patterns across different AI coding tools.

The Shift

From ad-hoc prompts to engineered autonomy

Good engineering teams already enforce planning, testing, code review, and scope management. But when teams adopt AI agents, those practices don't transfer automatically. AEF encodes them into the agent's workflow.

Agents without structure

  • Prompts are often written from scratch each time
  • Error recovery typically requires manual intervention
  • Output quality varies across different prompt styles
  • Context does not automatically carry between tasks

Explore

Pick your path

Understand the framework

Philosophy, principles, and the core concepts behind agentic engineering.

Design your workflow

Phases, prompt templates, quality gates, and how they connect.

Start building today

Practical guides to adopt the framework at your pace and team size.

The code is output. The agentic layer is the product.