The next is Half 3 of three from Addy Osmani’s authentic publish “Context Engineering: Bringing Engineering Self-discipline to Elements.” Half 1 may be discovered right here and Half 2 right here.
Context engineering is essential, but it surely’s only one element of a bigger stack wanted to construct full-fledged LLM purposes—alongside issues like management circulate, mannequin orchestration, software integration, and guardrails.
In Andrej Karpathy’s phrases, context engineering is “one small piece of an rising thick layer of non-trivial software program” that powers actual LLM apps. So whereas we’ve centered on learn how to craft good context, it’s essential to see the place that matches within the general structure.
A production-grade LLM system sometimes has to deal with many issues past simply prompting. For instance:
- Downside decomposition and management circulate: As an alternative of treating a consumer question as one monolithic immediate, sturdy programs typically break the issue down into subtasks or multistep workflows. As an example, an AI agent would possibly first be prompted to stipulate a plan, then in subsequent steps be prompted to execute every step. Designing this circulate (which prompts to name in what order; learn how to resolve branching or looping) is a basic programming job—besides the “capabilities” are LLM calls with context. Context engineering matches right here by ensuring every step’s immediate has the data it wants, however the resolution to have steps in any respect is a higher-level design. That is why you see frameworks the place you basically write a script that coordinates a number of LLM calls and power makes use of.
- Mannequin choice and routing: You would possibly use totally different AI fashions for various jobs. Maybe a light-weight mannequin for easy duties or preliminary solutions, and a heavyweight mannequin for closing options. Or a code-specialized mannequin for coding duties versus a normal mannequin for conversational duties. The system wants logic to route requests to the suitable mannequin. Every mannequin might need totally different context size limits or formatting necessities, which the context engineering should account for (e.g., truncating context extra aggressively for a smaller mannequin). This facet is extra engineering than prompting: consider it as matching the software to the job.
- Instrument integrations and exterior actions: In case your AI can carry out actions (like calling an API, database queries, opening an online web page, working code), your software program must handle these capabilities. That features offering the AI with an inventory of obtainable instruments and directions on utilization, in addition to really executing these software calls and capturing the outcomes. As we mentioned, the outcomes then grow to be new context for additional mannequin calls. Architecturally, this implies your app typically has a loop: immediate mannequin → if mannequin output signifies a software to make use of → execute software → incorporate outcome → immediate mannequin once more. Designing that loop reliably is a problem.
- Consumer interplay and UX flows: Many LLM purposes contain the consumer within the loop. For instance, a coding assistant would possibly suggest modifications after which ask the consumer to verify making use of them. Or a writing assistant would possibly provide a number of draft choices for the consumer to choose from. These UX choices have an effect on context too. If the consumer says “Choice 2 appears good however shorten it,” it is advisable carry that suggestions into the following immediate (e.g., “The consumer selected draft 2 and requested to shorten it.”). Designing a easy human-AI interplay circulate is a part of the app, although in a roundabout way about prompts. Nonetheless, context engineering helps it by making certain every flip’s immediate precisely displays the state of the interplay (like remembering which possibility was chosen or what the consumer edited manually).
- Guardrails and security: In manufacturing, it’s a must to think about misuse and errors. This would possibly embrace content material filters (to stop poisonous or delicate outputs), authentication and permission checks for instruments (so the AI doesn’t, say, delete a database as a result of it was within the directions), and validation of outputs. Some setups use a second mannequin or guidelines to double-check the primary mannequin’s output. For instance, after the principle mannequin generates a solution, you would possibly run one other test: “Does this reply comprise any delicate information? In that case, redact it.” These checks themselves may be carried out as prompts or as code. In both case, they typically add extra directions into the context (a system message like “If the consumer asks for disallowed content material, refuse,” is a part of many deployed prompts). So the context would possibly at all times embrace some security boilerplate. Balancing that (making certain the mannequin follows coverage with out compromising helpfulness) is yet one more piece of the puzzle.
- Analysis and monitoring: Suffice to say, it is advisable always monitor how the AI is performing. Logging each request and response (with consumer consent and privateness in thoughts) permits you to analyze failures and outliers. You would possibly incorporate real-time evals—e.g., scoring the mannequin’s solutions on sure standards, and if the rating is low, routinely having the mannequin attempt once more or path to a human fallback. Whereas analysis isn’t a part of producing a single immediate’s content material, it feeds again into bettering prompts and context methods over time. Primarily, you deal with the immediate and context meeting as one thing that may be debugged and optimized utilizing knowledge from manufacturing.
We’re actually speaking about a brand new form of utility structure. It’s one the place the core logic includes managing data (context) and adapting it via a sequence of AI interactions, moderately than simply working deterministic capabilities. Karpathy listed parts like management flows, mannequin dispatch, reminiscence administration, software use, verification steps, and many others., on high of context filling. All collectively, they kind what he jokingly calls “an rising thick layer” for AI apps—thick as a result of it’s doing lots! Once we construct these programs, we’re basically writing metaprograms: packages that choreograph one other “program” (the AI’s output) to resolve a job.
For us software program engineers, that is each thrilling and difficult. It’s thrilling as a result of it opens capabilities we didn’t have—e.g., constructing an assistant that may deal with pure language, code, and exterior actions seamlessly. It’s difficult as a result of most of the methods are new and nonetheless in flux. Now we have to consider issues like immediate versioning, AI reliability, and moral output filtering, which weren’t normal elements of app growth earlier than. On this context, context engineering lies on the coronary heart of the system: In case you can’t get the best data into the mannequin on the proper time, nothing else will save your app. However as we see, even good context alone isn’t sufficient; you want all of the supporting construction round it.
The takeaway is that we’re shifting from immediate design to system design. Context engineering is a core a part of that system design, but it surely lives alongside many different parts.
Conclusion
Key takeaway: By mastering the meeting of full context (and coupling it with strong testing), we will improve the possibilities of getting the most effective output from AI fashions.
For knowledgeable engineers, a lot of this paradigm is acquainted at its core—it’s about good software program practices—however utilized in a brand new area. Give it some thought:
- We at all times knew rubbish in, rubbish out. Now that precept manifests as “unhealthy context in, unhealthy reply out.” So we put extra work into making certain high quality enter (context) moderately than hoping the mannequin will determine it out.
- We worth modularity and abstraction in code. Now we’re successfully abstracting duties to a excessive stage (describe the duty, give examples, let AI implement) and constructing modular pipelines of AI + instruments. We’re orchestrating parts (some deterministic, some AI) moderately than writing all logic ourselves.
- We apply testing and iteration in conventional dev. Now we’re making use of the identical rigor to AI behaviors, writing evals and refining prompts as one would refine code after profiling.
In embracing context engineering, you’re basically saying, “I, the developer, am accountable for what the AI does.” It’s not a mysterious oracle; it’s a element I have to configure and drive with the best knowledge and guidelines.
This mindset shift is empowering. It means we don’t should deal with the AI as unpredictable magic—we will tame it with strong engineering methods (plus a little bit of artistic immediate artistry).
Virtually, how will you undertake this context-centric strategy in your work?
- Put money into knowledge and data pipelines. A giant a part of context engineering is having the info to inject. So construct that vector search index of your documentation, or arrange that database question that your agent can use. Deal with data sources as core options in growth. For instance, in case your AI assistant is for coding, ensure it may possibly pull in code from the repo or reference the model information. Loads of the worth you’ll get from an AI comes from the exterior data you provide to it.
- Develop immediate templates and libraries. Moderately than advert hoc prompts, begin creating structured templates on your wants. You might need a template for “reply with quotation” or “generate code diff given error.” These grow to be like capabilities you reuse. Preserve them in model management. Doc their anticipated habits. That is the way you construct up a toolkit of confirmed context setups. Over time, your workforce can share and iterate on these, simply as they’d on shared code libraries.
- Use instruments and frameworks that provide you with management. Keep away from “simply give us a immediate, we do the remaining” options in case you want reliability. Go for frameworks that allow you to peek beneath the hood and tweak issues—whether or not that’s a lower-level library like LangChain or a customized orchestration you construct. The extra visibility and management you may have over context meeting, the simpler debugging will likely be when one thing goes unsuitable.
- Monitor and instrument every thing. In manufacturing, log the inputs and outputs (inside privateness limits) so you possibly can later analyze them. Use observability instruments (like LangSmith, and many others.) to hint how context was constructed for every request. When an output is unhealthy, hint again and see what the mannequin noticed—was one thing lacking? Was one thing formatted poorly? This may information your fixes. Primarily, deal with your AI system as a considerably unpredictable service that it is advisable monitor like some other—dashboards for immediate utilization, success charges, and many others.
- Preserve the consumer within the loop. Context engineering isn’t nearly machine-machine information; it’s finally about fixing a consumer’s drawback. Usually, the consumer can present context if requested the best manner. Take into consideration UX designs the place the AI asks clarifying questions or the place the consumer can present further particulars to refine the context (like attaching a file, or choosing which codebase part is related). The time period “AI-assisted” goes each methods—AI assists the consumer, however the consumer can help AI by supplying context. A well-designed system facilitates that. For instance, if an AI reply is unsuitable, let the consumer appropriate it and feed that correction again into context for subsequent time.
- Prepare your workforce (and your self). Make context engineering a shared self-discipline. In code opinions, begin reviewing prompts and context logic too. (“Is that this retrieval grabbing the best docs? Is that this immediate part clear and unambiguous?”) In case you’re a tech lead, encourage workforce members to floor points with AI outputs and brainstorm how tweaking context would possibly repair it. Information sharing is vital as a result of the sphere is new—a intelligent immediate trick or formatting perception one particular person discovers can doubtless profit others. I’ve personally realized a ton simply studying others’ immediate examples and postmortems of AI failures.
As we transfer ahead, I anticipate context engineering to grow to be second nature—very like writing an API name or a SQL question is right this moment. It will likely be a part of the usual repertoire of software program growth. Already, many people don’t assume twice about doing a fast vector similarity search to seize context for a query; it’s simply a part of the circulate. In a number of years, “Have you ever arrange the context correctly?” will likely be as frequent a code evaluate query as “Have you ever dealt with that API response correctly?”
In embracing this new paradigm, we don’t abandon the outdated engineering ideas—we reapply them in new methods. In case you’ve spent years honing your software program craft, that have is extremely worthwhile now: It’s what permits you to design wise flows, spot edge instances, and guarantee correctness. AI hasn’t made these abilities out of date; it’s amplified their significance in guiding AI. The function of the software program engineer shouldn’t be diminishing—it’s evolving. We’re changing into administrators and editors of AI, not simply writers of code. And context engineering is the approach by which we direct the AI successfully.
Begin pondering by way of what data you present to the mannequin, not simply what query you ask. Experiment with it, iterate on it, and share your findings. By doing so, you’ll not solely get higher outcomes from right this moment’s AI but additionally be making ready your self for the much more highly effective AI programs on the horizon. Those that perceive learn how to feed the AI will at all times have the benefit.
Completely satisfied context-coding!
I’m excited to share that I’ve written a brand new AI-assisted engineering e-book with O’Reilly. In case you’ve loved my writing right here you might be enthusiastic about checking it out.
AI instruments are rapidly shifting past chat UX to stylish agent interactions. Our upcoming AI Codecon occasion, Coding for the Agentic World, will spotlight how builders are already utilizing brokers to construct progressive and efficient AI-powered experiences. We hope you’ll be a part of us on September 9 to discover the instruments, workflows, and architectures defining the following period of programming. It’s free to attend. Register now to save lots of your seat.
