⌘K
Change language Switch ThemeSign In
Narrow Mode
What is agentic engineering? - Agentic Engineering Patterns ===========================================================
S Simon Willison's Weblog @Simon Willison
One Sentence Summary
The article defines agentic engineering as goal-driven tool loops by coding agents and explains why human judgment remains central to software development.
Summary
This guide introduces “agentic engineering” as a development practice where coding agents write and execute code iteratively to reach a user-defined goal. Its strongest contribution is a practical definition of LLM agents in software work: tool use in a loop, with executable feedback as the core differentiator from plain text generation. The article also reframes the human role as problem framing, tradeoff evaluation, tool harness design, and iterative verification. While it is an introductory chapter rather than a deep technical benchmark, it offers a clear conceptual foundation for teams adopting coding agents in real workflows.
Main Points
* 1. Agentic engineering is defined by executable tool loops, not one-shot code generation.The article argues that the ability to run code and iterate against outcomes is what makes coding agents practically useful. * 2. Human value shifts from typing code to framing and evaluating solutions.Engineers remain responsible for selecting goals, defining constraints, and making tradeoff decisions across candidate implementations. * 3. Instruction and tool harness design determine long-term agent performance.The piece highlights that agent quality improves when teams continuously refine prompts, tools, and verification steps from observed failures.
Metadata
AI Score
87
Website simonwillison.net
Published At Yesterday
Length 473 words (about 2 min)
Sign in to use highlight and note-taking features for a better reading experience. Sign in now Guides>Agentic Engineering Patterns
I use the term agentic engineering to describe the practice of developing software with the assistance of coding agents.
What are coding agents? They're agents that can both write and execute code. Popular examples include Claude Code, OpenAI Codex, and Gemini CLI.
What's an agent? Clearly defining that term is a challenge that has frustrated AI researchers since at least the 1990s but the definition I've come to accept, at least in the field of Large Language Models (LLMs) like GPT-5 and Gemini and Claude, is this one: Agents run tools in a loop to achieve a goal You prompt the coding agent to define a goal. The agent then generates and executes code in a loop until that goal has been met.
Code execution is the defining capability that makes agentic engineering possible. Without the ability to directly run the code, anything output by an LLM is of limited value. With code execution, these agents can start iterating towards software that demonstrably works.
Agentic engineering -------------------
Now that we have software that can write working code, what is there left for us humans to do?
The answer is _so much stuff_.
Writing code has never been the sole activity of a software engineer. The craft has always been figuring out _what_ code to write. Any given software problem has dozens of potential solutions, each with their own tradeoffs. Our job is to navigate those options and find the ones that are the best fit for our unique set of circumstances and requirements.
Getting great results out of coding agents is a deep subject in its own right, especially now as the field continues to evolve at a bewildering rate.
We need to provide our coding agents with the tools they need to solve our problems, specify those problems in the right level of detail, and verify and iterate on the results until we are confident they address our problems in a robust and credible way.
LLMs don't learn from their past mistakes, but coding agents can, provided we deliberately update our instructions and tool harnesses to account for what we learn along the way.
Used effectively, coding agents can help us be much more ambitious with the projects we take on. Agentic engineering should help us produce more, better quality code that solves more impactful problems.
About this guide ----------------
Just like the field it attempts to cover, _Agentic Engineering Patterns_ is very much a work in progress. My goal is to identify and describe patterns for working with these tools that demonstrably get results, and that are unlikely to become outdated as the tools advance.
I'll continue adding more chapters as new techniques emerge. No chapter should be considered finished. I'll be updating existing chapters as our understanding of these patterns evolves. Writing code is cheap now →
S Simon Willison's Weblog @Simon Willison
One Sentence Summary
The article defines agentic engineering as goal-driven tool loops by coding agents and explains why human judgment remains central to software development.
Summary
This guide introduces “agentic engineering” as a development practice where coding agents write and execute code iteratively to reach a user-defined goal. Its strongest contribution is a practical definition of LLM agents in software work: tool use in a loop, with executable feedback as the core differentiator from plain text generation. The article also reframes the human role as problem framing, tradeoff evaluation, tool harness design, and iterative verification. While it is an introductory chapter rather than a deep technical benchmark, it offers a clear conceptual foundation for teams adopting coding agents in real workflows.
Main Points
* 1. Agentic engineering is defined by executable tool loops, not one-shot code generation.
The article argues that the ability to run code and iterate against outcomes is what makes coding agents practically useful.
* 2. Human value shifts from typing code to framing and evaluating solutions.
Engineers remain responsible for selecting goals, defining constraints, and making tradeoff decisions across candidate implementations.
* 3. Instruction and tool harness design determine long-term agent performance.
The piece highlights that agent quality improves when teams continuously refine prompts, tools, and verification steps from observed failures.
Key Quotes
* Agents run tools in a loop to achieve a goal * Code execution is the defining capability that makes agentic engineering possible. * Writing code has never been the sole activity of a software engineer.
AI Score
87
Website simonwillison.net
Published At Yesterday
Length 473 words (about 2 min)
Tags
Agentic Engineering
Coding Agents
Tool Loop
LLM Workflow
Human-in-the-Loop
Related Articles
* First impressions of Claude Cowork， Anthropic’s general agent * How we built LangChain’s GTM Agent * OpenClaw: The Viral AI Agent that Broke the Internet - Peter Steinberger | Lex Fridman Podcast #491 * 2025: The year in LLMs * How StrongDM’s AI team build serious software without even looking at the code * My fireside chat about agentic engineering at the Pragmatic Summit * Wilson Lin on FastRender: a browser built by thousands of parallel agents * Design-First Collaboration * How Claude Code Works - Jared Zoneraich， PromptLayer * Introducing Showboat and Rodney， so agents can demo what they’ve built HomeArticlesPodcastsVideosTweets
What is agentic engineering? - Agentic Engineering Patter... ===============