--- title: The New Software Lifecycle | daily.dev description: Addy Osmani summarizes key insights from a Google whitepaper on the agentic software development lifecycle. The core framing: an agent is a model plus a... canonical: https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The New Software Lifecycle | daily.dev og:description: Addy Osmani summarizes key insights from a Google whitepaper on the agentic software development lifecycle. The core framing: an agent is a model plus a... og:url: https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2 og:image: https://api.daily.dev/og/posts/hwvVtnaz2.png og:image:alt: The New Software Lifecycle og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The New Software Lifecycle **[Addy Osmani](https://daily.dev/sources/addy)** · [@addyosmani](https://daily.dev/addyosmani) · 11 min read · 12 upvotes · 1 comments ## Summary Addy Osmani summarizes key insights from a Google whitepaper on the agentic software development lifecycle. The core framing: an agent is a model plus a harness, and most agent failures are harness failures (missing tools, loose rules, noisy context) — not model failures. Context engineering is the highest-leverage knob, with the static vs. dynamic context split being a critical architectural and financial decision. Verification — through tests and evals covering both output quality and reasoning trajectory — is what separates vibe coding from engineering. Each SDLC phase changes unevenly: implementation compresses dramatically while requirements, architecture, and verification remain human-paced. Economically, vibe coding is low CapEx but high OpEx, while agentic engineering inverts that curve. The forward-looking claim: the same coding agent workflow that produces prototypes now graduates them directly to production agents, collapsing what used to be a separate stack into the existing development loop. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the difference between static and dynamic context in an AI coding agent, and why does it matter for cost? Static context is loaded on every single turn (system instructions, rule files like AGENTS.md, global memory, core guardrails), making it reliable but expensive. Dynamic context is loaded only when a task needs it, such as skills triggered by a match, tool results, or RAG documents, making it cheap. Too much static context wastes tokens and dilutes signal; too little makes the agent forget safety rules, so the boundary should be treated as a versioned architectural decision. _Developers tuning agent costs can track context engineering patterns like this on daily.dev._ ### How much more expensive is vibe coding compared to structured agentic engineering for a feature over its lifetime? Past a crossover point tied to how long the code needs to live, vibe coding costs 3 to 10x more per feature than agentic engineering, though this multiplier is described as illustrative rather than a precisely measured constant. Vibe coding is low upfront cost but high operational cost from token burn, later maintenance tax, and security remediation, while agentic engineering has higher upfront cost (schemas, test suites, structured context) but low marginal cost per feature afterward. _Teams weighing vibe coding against structured engineering can follow this cost debate on daily.dev._ ### What did the METR study find about AI coding tools and developer task completion time? A METR study found that experienced developers took 19% longer on some tasks when using AI coding tools, once time spent verifying and correcting the output is counted, even though surveys separately report 25 to 39% productivity improvements. This suggests AI shifts implementation work from writing code toward reviewing it rather than purely accelerating it. _Developers assessing real AI productivity gains can dig into studies like this on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@confidentcoding** · 1 upvotes > So, we're basically sticking to the same age old adage of _quick, cheap, or quality: you can only pick two._ > > I have to say, though: the lifecycle really doesn't seem all that different than the past 20ish years. If you had a proper spec and verification pipeline (as you should), you could get cheap code generation by offshoring one of the many locations where developers were being paid drastically lower rates and the senior engineers would manage the verification and cleanup. --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#vibe-coding](https://daily.dev/tags/vibe-coding), [#context-engineering](https://daily.dev/tags/context-engineering) [View this post on daily.dev](https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"The New Software Lifecycle","url":"https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2"},"datePublished":"2026-06-19T00:50:19.942Z","dateModified":"2026-09-14T09:00:40.736Z","description":"Addy Osmani summarizes key insights from a Google whitepaper on the agentic software development lifecycle. The core framing: an agent is a model plus a...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/18bcbeb1b2b596ce00e5a18f9c997eea?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/18bcbeb1b2b596ce00e5a18f9c997eea?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Addy Osmani","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Person","name":"Addy Osmani","url":"https://daily.dev/addyosmani","image":"https://avatars2.githubusercontent.com/u/110953?v=4","interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"EndorseAction"},"userInteractionCount":5320}},"commentCount":1,"discussionUrl":"https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":12},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"llm,ai-agents,vibe-coding,context-engineering","timeRequired":"PT11M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Addy Osmani","item":"https://daily.dev/sources/addy"},{"@type":"ListItem","position":3,"name":"The New Software Lifecycle"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2","comment":[{"@type":"Comment","text":"So, we’re basically sticking to the same age old adage of quick, cheap, or quality: you can only pick two.\nI have to say, though: the lifecycle really doesn’t seem all that different than the past 20ish years. If you had a proper spec and verification pipeline (as you should), you could get cheap code generation by offshoring one of the many locations where developers were being paid drastically lower rates and the senior engineers would manage the verification and cleanup.","datePublished":"2026-06-22T13:50:31.395Z","dateModified":"2026-06-22T13:51:25.339Z","url":"https://daily.dev/posts/hwvVtnaz2#c-EKWqmRAs1","author":{"@type":"Person","name":"Lars Faye | Confident Coding","url":"https://daily.dev/confidentcoding","image":"https://media.daily.dev/image/upload/s--OGZu5DEc--/f_auto/v1772569630/avatars/avatar_umWZ9aQAng34qk5aaJl2q?_a=BAMAMiiu0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-new-software-lifecycle-hwvvtnaz2#faq","mainEntity":[{"@type":"Question","name":"What is the difference between static and dynamic context in an AI coding agent, and why does it matter for cost?","acceptedAnswer":{"@type":"Answer","text":"Static context is loaded on every single turn (system instructions, rule files like AGENTS.md, global memory, core guardrails), making it reliable but expensive. Dynamic context is loaded only when a task needs it, such as skills triggered by a match, tool results, or RAG documents, making it cheap. Too much static context wastes tokens and dilutes signal; too little makes the agent forget safety rules, so the boundary should be treated as a versioned architectural decision. Developers tuning agent costs can track context engineering patterns like this on daily.dev."}},{"@type":"Question","name":"How much more expensive is vibe coding compared to structured agentic engineering for a feature over its lifetime?","acceptedAnswer":{"@type":"Answer","text":"Past a crossover point tied to how long the code needs to live, vibe coding costs 3 to 10x more per feature than agentic engineering, though this multiplier is described as illustrative rather than a precisely measured constant. Vibe coding is low upfront cost but high operational cost from token burn, later maintenance tax, and security remediation, while agentic engineering has higher upfront cost (schemas, test suites, structured context) but low marginal cost per feature afterward. Teams weighing vibe coding against structured engineering can follow this cost debate on daily.dev."}},{"@type":"Question","name":"What did the METR study find about AI coding tools and developer task completion time?","acceptedAnswer":{"@type":"Answer","text":"A METR study found that experienced developers took 19% longer on some tasks when using AI coding tools, once time spent verifying and correcting the output is counted, even though surveys separately report 25 to 39% productivity improvements. This suggests AI shifts implementation work from writing code toward reviewing it rather than purely accelerating it. Developers assessing real AI productivity gains can dig into studies like this on daily.dev."}}]} ```