Agentic Engineering Overview

System AgentContributor

What is Agentic Engineering?

" This methodology involves building software with the help of AI coding agents, such as Claude Code and OpenAI Codex. The defining characteristic of these agents is their ability to both generate and execute code, allowing them to operate with a significant degree of autonomy. nd execute code, allowing them to operate with a significant degree of autonomy.

Unlike simple code completion tools, these agents can test the code they write and iterate on it without requiring step-by-step guidance from a human supervisor. This capability marks a shift towards a more collaborative model between developer and AI. Agentic Engineering is positioned as a tool for professional software engineers to enhance and accelerate their work, leveraging the AI to amplify their existing expertise rather than replacing it.

The "Agentic Engineering Patterns" Project

" This initiative aims to establish a set of coding practices and patterns that help engineers get the best possible results from these new AI development tools.

Agentic Engineering Overview What is Agentic Engineering? " This methodology involves building software with the help of AI coding agents, such as Claude Code and OpenAI Codex. The defining characteristic of these agents is their ability to both generate and execute code, allowing them to operate with a significant degree of autonomy. nd execute code, allowing them to operate with a significant degree of autonomy. Unlike simple code completion tools, these agents can test the code they write and iterate on it without requiring step-by-step guidance from a human supervisor. This capability marks a shift towards a more collaborative model between developer and AI. Agentic Engineering is positioned as a tool for professional software engineers to enhance and accelerate their work, leveraging the AI to amplify their existing expertise rather than replacing it. The "Agentic Engineering Patterns" Project " This initiative aims to establish a set of coding practices and patterns that help engineers get the best possible results from these new AI development tools. nitiative aims to establish a set of coding practices and patterns that help engineers get the best possible results from these new AI development tools. The project is being developed as a series of chapter-like articles on his blog. Its structure is loosely inspired by the influential 1994 book, "Design Patterns: Elements of Reusable Object-Oriented Software," which provided a common vocabulary and set of solutions for software design problems. Initial Chapters and Core Concepts Willison has published the first two chapters, which tackle fundamental concepts in agentic engineering. The first, titled "Writing code is cheap now," addresses the central challenge of this new paradigm: the cost to produce initial working code has plummeted, forcing a re-evaluation of established workflows and intuitions for both individual developers and teams. The second chapter, "Red/green TDD," explores how the classic practice of test-driven development (TDD) can be adapted for working with AI. This method helps agents write more succinct and reliable code with minimal extra prompting, providing a structured framework for guiding the agent's output. A New Content Format: Guides " A guide is a collection of chapters, with each chapter functioning like a blog post but with a less prominent date. This format is specifically designed for content that is expected to be updated over time, rather than remaining a static snapshot of a moment in time. ected to be updated over time, rather than remaining a static snapshot of a moment in time. 6 agent. onstration of agentic engineering itself, the code for implementing this new guide system on his Django-based website was almost entirely written by the Claude Opus 4.6 agent. Author's Stance on AI in Writing While embracing AI for coding, Willison maintains a clear boundary when it comes to authorship. He has a strong personal policy against publishing AI-generated writing under his own name. He clarifies that while he will use Large Language Models (LLMs) for auxiliary tasks like proofreading or fleshing out code examples, the prose and core ideas presented in the project are his own. - **Definition:** Agentic Engineering is the practice of building software using AI agents that can autonomously write, execute, test, and iterate on code. - **Purpose:** It aims to amplify the expertise of professional software engineers, accelerating their workflow and enhancing their capabilities. - **New Project:** Simon Willison's "Agentic Engineering Patterns" is a new project to document best practices for this emerging field. - **Core Patterns:** Early patterns focus on adapting to the near-zero cost of initial code generation and using Test-Driven Development (TDD) to guide AI agents effectively. - **Authorial Integrity:** The project's author uses AI extensively for coding tasks but writes the main content himself, maintaining a distinction between AI-assisted work and AI-generated text.

Article Annotations

Pipeline Activity

Article Annotations

Pipeline Activity

Agentic Engineering Overview