--- title: "The Agentic Codebase — STACK Architecture for AI Repos | Len P. van der Hof" description: "Agents do not fail only because models are weak. They fail because repos were never designed for machine collaborators. The Agentic Codebase gives you STACK: Structure, Toolchain, Agent configuration, Connection layer, Knowledge and quality." image: "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.wide.webp" --- [AI Systems](https://lenvanderhof.com/en/blog/category/ai-systems/) Research guide # The Agentic Codebase: Version the Agent OS Like You Version the App Prompts in chat are not architecture. Scattered rules are not governance. STACK turns your repository into a maintainable agent operating system. Len P. van der HofPublished 5 August 2026Updated 5 August 20263 min read Production-grade architecture for repositories where agents work. Direct answer The Agentic Codebase is a production architecture book for technical founders and senior builders shipping agentic software. It introduces the STACK five-layer model so AGENTS.md, rules, skills, MCP, hooks, evals, and shell automation become versioned repository infrastructure that survives tool churn and team growth. ## Key takeaways - Agent failure is often repository debt — scattered rules, unowned skills, MCP without contracts — not model weakness. - STACK unifies structure, toolchain, agent config, connections, and quality into one maintainable system. - Version the agent OS like the app: coherent instructions, fail-closed integrations, hooks, and evals that catch regressions. ## You can demo an agent in an afternoon. You can regret it in six months. Prompts live in chat history. Rules are copy-pasted across five tools. MCP servers sprawl without contracts. Skills exist but nobody knows which agent loads them. “Vibe coding” ships until the repository becomes unnavigable for humans **and** machines. That is not a model problem first. That is **agentic repo debt**. *The Agentic Codebase* names the disease and installs the cure. ## The thesis that cuts through tool noise Agents do not fail only because models are weak. They fail because repositories, rules, and tool contracts were never designed for machine collaborators. So stop treating the agent surface as disposable chat. **Version the agent OS like you version the app.** ## STACK: five layers, one maintainable system STACK is the architecture: 1. **Structure** — the repo tree agents and humans can both navigate. 2. **Toolchain** — shells, runners, and deterministic workflows that do not rot. 3. **Agent configuration** — AGENTS.md / CLAUDE.md, rules, skills, roles, and charters that agree. 4. **Connection layer** — MCP and integrations with contracts, not hope. 5. **Knowledge and quality** — memory, evals, hooks, CI gates, and regression discipline. Five layers. One operating system for agent work. ## What you can ship after reading Not vibes. Artifacts. - a coherent agent instruction surface across tools - skill packages that are discoverable and owned - MCP tool contracts that fail closed when wrong - hooks and evals that catch regressions before users do - a STACK audit you can run on a brownfield repo this week This is software engineering applied to the agent instruction plane. ## Who this book is for - the staff engineer whose team uses Claude Code three incompatible ways - the founder shipping agent features into a real product - the platform lead standardizing AI coding practice - the maintainer who inherited `CLAUDE.md`, `.cursorrules`, and `AGENTS.md` that disagree - the operator who already read the strategy book and now needs the implementation layer If you want a one-tool tour, this will feel rigid. If you want production survival, that rigidity is the product. ## Series context without soft cross-sell fog - *AI Agents for Startup Strategy* answers **where** agents earn a place. - *The Agentic Author* shows gated production for knowledge work. - *The Agentic Codebase* hardens the **repository** those agents live in. Different layer. Same seriousness. ## Buy the edition your team will actually use **[Open the book page →](https://lenvanderhof.com/books/the-agentic-codebase/)** ### English editions (Amazon.com) FormatLinkKindleBuy English KindlePaperbackBuy English paperbackHardcoverBuy English hardcover ### Dutch editions (Amazon.nl) FormatLinkKindleKoop Nederlandse KindlePaperbackKoop Nederlandse paperbackHardcoverKoop Nederlandse hardcover Dutch page: [boeken/the-agentic-codebase](https://lenvanderhof.com/boeken/the-agentic-codebase/). ## Close the loop Vibe coding is real. Without a lane, it is a path to repo rot. Name the lane. Instrument the gate. Keep the agent OS versioned. ## Frequently asked questions Is The Agentic Codebase a Cursor or Claude Code tutorial? No. It is cross-tool architecture for agentic repositories. Tool names change; STACK principles are designed to outlast them. What does STACK stand for? Structure, Toolchain, Agent configuration, Connection layer, Knowledge and quality — five layers every agentic codebase must get right. Who should read this book? Technical founders, staff engineers, engineering managers, and maintainers who need agent work to survive production, not just demo day. How does this relate to AI Agents for Startup Strategy? Strategy decides where to deploy agents. The Agentic Codebase is the implementation layer: how the repository runs them safely and repeatably. Which editions can I buy? English and Dutch Kindle, paperback, and hardcover editions via Amazon, with full links on the book page and in the buy grid below. ## Sources 1. [The Agentic Codebase — book page](https://lenvanderhof.com/books/the-agentic-codebase/) · Lenvanderhof.com 2. [English Kindle listing](https://www.amazon.com/dp/B0H964XXTS) · Amazon ## Related essays - [What is an AI agent?](https://lenvanderhof.com/en/blog/what-is-an-ai-agent/) - [MCP explained for founders](https://lenvanderhof.com/en/blog/mcp-explained-for-founders/) - [Version the agent OS](https://lenvanderhof.com/en/blog/stack-version-the-agent-os/) - [An agentic workflow is not an automation](https://lenvanderhof.com/en/blog/agentic-workflow-vs-automation/) - [LLM integration is an operating-design problem](https://lenvanderhof.com/en/blog/llm-integration-is-operating-design/) - [MCP meaning: Model Context Protocol, not a product](https://lenvanderhof.com/en/blog/mcp-meaning/) - [What is STACK? Five layers, not a vendor diagram](https://lenvanderhof.com/en/blog/what-is-stack/) ## Further reading - [Full book page and buy grid](https://lenvanderhof.com/books/the-agentic-codebase/) - [STACK framework](https://lenvanderhof.com/frameworks/stack/) About the author ## [Len P. van der Hof](https://lenvanderhof.com/en/authors/len-p-van-der-hof/) Entrepreneur, AI Innovator and Venture Builder Len P. van der Hof builds practical AI systems, digital ventures and evidence-informed tools for founders. ```json { "@context": "https://schema.org", "@graph": [ { "@type": "Person", "@id": "https://lenvanderhof.com/#person", "name": "Len P. van der Hof", "alternateName": [ "Len van der Hof", "L.P. van der Hof", "Leendert Pieter van der Hof" ], "honorificSuffix": "MSc", "url": "https://lenvanderhof.com/", "image": [ "https://lenvanderhof.com/photos/len-portrait-1.jpg", "https://lenvanderhof.com/photos/len-portrait-2.jpg", "https://lenvanderhof.com/photos/len-portrait-3.jpg", "https://lenvanderhof.com/photos/len-portrait-4.jpg", "https://lenvanderhof.com/photos/len-speaking.jpg", "https://lenvanderhof.com/photos/len-hero.jpg" ], "jobTitle": "Entrepreneur, AI Innovator and Venture Builder", "description": "Len P. van der Hof is a Dutch entrepreneur and AI innovator building ReasonKit, MindSesh, Undominated.ai, the fiction imprint LPH98.lifestyle, and technology ventures through LPH98.ventures.", "address": { "@type": "PostalAddress", "addressLocality": "Zwijndrecht", "addressCountry": "NL" }, "alumniOf": { "@type": "CollegeOrUniversity", "name": "Rotterdam School of Management, Erasmus University" }, "knowsAbout": [ "Artificial intelligence", "AI agents", "Agentic AI systems", "LLM routing", "SEO", "Generative engine optimization", "Answer engine optimization", "Venture building", "Founder performance", "Founder psychology", "Evidence-based decision-making" ], "sameAs": [ "https://www.linkedin.com/in/lenvanderhof/", "https://x.com/LenvanderHof", "https://www.youtube.com/channel/UCTG20buKqYYbitqqf7l3zJA", "https://www.instagram.com/Lenvanderhof/", "https://www.threads.com/@lenvanderhof", "https://github.com/Lenvanderhof", "https://huggingface.co/LPH98", "https://www.npmjs.com/~lenvanderhof", "https://www.goodreads.com/author/show/70983905.Len_P_van_der_Hof", "https://www.amazon.com/author/lenvanderhof", "https://www.bol.com/nl/nl/b/len-p-van-der-hof-msc/609879394/", "https://bsky.app/profile/lenvanderhof.com", "https://mastodon.social/@Lenvanderhof", "https://crates.io/users/Lenvanderhof", "https://cursor.com/@Lenvanderhof", "https://medium.com/@Lenvanderhof", "https://gitlab.com/Lenvanderhof", "https://hub.docker.com/u/lenvanderhof/", "https://dev.to/lenvanderhof", "https://www.facebook.com/Lenvanderhof", "https://soundcloud.com/Lenvanderhof" ], "affiliation": [ { "@id": "https://lenvanderhof.com/#publisher" }, { "@id": "https://lenvanderhof.com/#mindsesh" }, { "@id": "https://lenvanderhof.com/#lifestyle" } ] }, { "@type": "WebSite", "@id": "https://lenvanderhof.com/#website", "url": "https://lenvanderhof.com/", "name": "Len P. van der Hof", "description": "Len P. van der Hof is a Dutch entrepreneur and AI innovator building ReasonKit, MindSesh, Undominated.ai, the fiction imprint LPH98.lifestyle, and technology ventures through LPH98.ventures.", "inLanguage": [ "en", "nl" ], "publisher": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#publisher", "name": "LPH98.ventures", "url": "https://lph98.ventures", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#mindsesh", "name": "MindSesh", "url": "https://mindsesh.net", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "SoftwareApplication", "@id": "https://lenvanderhof.com/#reasonkit", "name": "ReasonKit", "url": "https://reasonkit.sh", "creator": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "SoftwareApplication", "@id": "https://lenvanderhof.com/#undominated", "name": "Undominated.ai", "url": "https://undominated.ai", "creator": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "Organization", "@id": "https://lenvanderhof.com/#lifestyle", "name": "LPH98.lifestyle", "url": "https://lph98.lifestyle", "founder": { "@id": "https://lenvanderhof.com/#person" } }, { "@type": "ImageObject", "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#primaryimage", "url": "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.wide.webp", "contentUrl": "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.wide.webp", "representativeOfPage": true }, { "@type": "BreadcrumbList", "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#breadcrumb", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://lenvanderhof.com/" }, { "@type": "ListItem", "position": 2, "name": "Blog", "item": "https://lenvanderhof.com/en/blog/" }, { "@type": "ListItem", "position": 3, "name": "AI Systems", "item": "https://lenvanderhof.com/en/blog/category/ai-systems/" }, { "@type": "ListItem", "position": 4, "name": "The Agentic Codebase: Version the Agent OS Like You Version the App", "item": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/" } ] }, { "@type": "WebPage", "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#webpage", "url": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/", "name": "The Agentic Codebase: Version the Agent OS Like You Version the App", "description": "Agents do not fail only because models are weak. They fail because repos were never designed for machine collaborators. The Agentic Codebase gives you STACK: Structure, Toolchain, Agent configuration, Connection layer, Knowledge and quality.", "isPartOf": { "@id": "https://lenvanderhof.com/#website" }, "primaryImageOfPage": { "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#primaryimage" }, "breadcrumb": { "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#breadcrumb" }, "inLanguage": "en-GB" }, { "@type": "BlogPosting", "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#article", "mainEntityOfPage": { "@id": "https://lenvanderhof.com/en/blog/the-agentic-codebase-introduction/#webpage" }, "headline": "The Agentic Codebase: Version the Agent OS Like You Version the App", "description": "Agents do not fail only because models are weak. They fail because repos were never designed for machine collaborators. The Agentic Codebase gives you STACK: Structure, Toolchain, Agent configuration, Connection layer, Knowledge and quality.", "datePublished": "2026-08-05T13:00:00.000Z", "dateModified": "2026-08-05T17:00:00.000Z", "author": { "@id": "https://lenvanderhof.com/#person" }, "publisher": { "@id": "https://lenvanderhof.com/#person" }, "image": [ "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.square.webp", "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.landscape.webp", "https://lenvanderhof.com/media/generated/blog-hero-tac-en.c5f96e2cbf13.wide.webp" ], "articleSection": "AI Systems", "inLanguage": "en-GB" } ] } ```