Thoughts on AI-native development, agent workflows, design, and engineering.https://interlusion.dev/en/blog/enhttps://interlusion.dev/en/blog/thinking-in-a-borrowed-language/https://interlusion.dev/en/blog/thinking-in-a-borrowed-language/AI models reason in a language they borrowed from us, and training is quietly rewriting it. An essay on compounds, telegrams, scratch paper and handoffs.Wed, 28 Oct 2026 00:00:00 GMTphilosophyresearchhttps://interlusion.dev/en/blog/why-ai-models-think-on-scratch-paper/https://interlusion.dev/en/blog/why-ai-models-think-on-scratch-paper/Reasoning models write notes to themselves before they answer. Why they need this scratch paper, why it is getting harder to read, and why that matters.Tue, 27 Oct 2026 00:00:00 GMTresearchagentshttps://interlusion.dev/en/blog/they-rigged-me/https://interlusion.dev/en/blog/they-rigged-me/Darbit, the rabbit in Interlusion's logo, on being animated: bending ears, masked eyes, a hat that spins without moving, and one rule: stay the logo.Fri, 16 Oct 2026 00:00:00 GMTuiengineeringhttps://interlusion.dev/en/blog/google-stitch-vibe-design-and-the-design-md-pipeline/https://interlusion.dev/en/blog/google-stitch-vibe-design-and-the-design-md-pipeline/Google's AI design tool generates high-fidelity UI from prompts, sketches, and voice. The interesting part is DESIGN.md and its MCP bridge to your code.Mon, 30 Mar 2026 00:00:00 GMTuitoolsai-nativehttps://interlusion.dev/en/blog/paper-the-code-native-design-canvas/https://interlusion.dev/en/blog/paper-the-code-native-design-canvas/Paper eliminates the design-to-code translation by making the canvas itself HTML and CSS. What you design is what ships. The question is whether that's always what you want.Sun, 29 Mar 2026 00:00:00 GMTuitoolsai-nativehttps://interlusion.dev/en/blog/pencil-design-that-lives-where-the-code-lives/https://interlusion.dev/en/blog/pencil-design-that-lives-where-the-code-lives/Pencil puts a Figma-like canvas inside your IDE with design files in Git and AI agents reading specs through MCP. The design never leaves the codebase.Sat, 28 Mar 2026 00:00:00 GMTuitoolsai-nativehttps://interlusion.dev/en/blog/the-design-handoff-is-dying/https://interlusion.dev/en/blog/the-design-handoff-is-dying/Three new tools close the gap between design and code from different directions. The deeper shift: agents need a shared design vocabulary to work at all.Fri, 27 Mar 2026 00:00:00 GMTuiopinionai-nativehttps://interlusion.dev/en/blog/astro-the-framework-that-ships-less/https://interlusion.dev/en/blog/astro-the-framework-that-ships-less/Why Interlusion builds its marketing site with Astro, and what a zero-JavaScript-by-default architecture means for performance, simplicity, and AI-native development.Thu, 26 Mar 2026 00:00:00 GMTengineeringtoolsuihttps://interlusion.dev/en/blog/building-a-design-token-system-with-tailwind-v4/https://interlusion.dev/en/blog/building-a-design-token-system-with-tailwind-v4/A walkthrough of a three-layer design token architecture with Tailwind CSS v4's native @theme directive. No config files, no build tools, just CSS.Tue, 24 Mar 2026 00:00:00 GMTuiengineeringtoolshttps://interlusion.dev/en/blog/oklch-the-end-of-unpredictable-color/https://interlusion.dev/en/blog/oklch-the-end-of-unpredictable-color/HSL promised human-readable colour and lied about lightness. OKLCH delivers: perceptually uniform colour for P3 displays, design tokens, and AI agents.Tue, 24 Mar 2026 00:00:00 GMTuiengineeringhttps://interlusion.dev/en/blog/design-tokens-the-api-between-design-and-code/https://interlusion.dev/en/blog/design-tokens-the-api-between-design-and-code/Design systems fail when design and code drift apart. Design tokens fix the interface: one source of truth that both sides can read, write, and ship from.Mon, 23 Mar 2026 00:00:00 GMTuiengineeringopinionhttps://interlusion.dev/en/blog/pageindex-and-the-case-against-chunking/https://interlusion.dev/en/blog/pageindex-and-the-case-against-chunking/VectifyAI's PageIndex swaps vector search for tree reasoning and scores 98.7% on FinanceBench. The result is real. The question is what it actually proves.Fri, 20 Mar 2026 00:00:00 GMTresearchengineeringraghttps://interlusion.dev/en/blog/most-rag-fails-at-the-r-not-the-g/https://interlusion.dev/en/blog/most-rag-fails-at-the-r-not-the-g/Everyone obsesses over models and prompts. The real bottleneck in RAG is retrieval, and fixing it takes engineering, not bigger context windows.Thu, 19 Mar 2026 00:00:00 GMTagentsengineeringai-nativeraghttps://interlusion.dev/en/blog/rag-101-what-it-is-and-why-it-matters/https://interlusion.dev/en/blog/rag-101-what-it-is-and-why-it-matters/A plain-language introduction to Retrieval-Augmented Generation: what it does, how it works, and why it is becoming the default way to use AI on your data.Thu, 19 Mar 2026 00:00:00 GMTai-nativeengineeringraghttps://interlusion.dev/en/blog/sqlc-the-tool-that-makes-sql-first-development-practical/https://interlusion.dev/en/blog/sqlc-the-tool-that-makes-sql-first-development-practical/sqlc generates type-safe Go code from plain SQL. No ORM, no runtime reflection, no surprises: your schema, your queries, and generated code that compiles.Wed, 18 Mar 2026 00:00:00 GMTengineeringgotoolshttps://interlusion.dev/en/blog/autoresearch-the-night-shift-that-changes-everything/https://interlusion.dev/en/blog/autoresearch-the-night-shift-that-changes-everything/Andrej Karpathy's autoresearch lets AI agents run about a hundred ML experiments overnight on one GPU. It looks like the beginning of autonomous science.Tue, 17 Mar 2026 00:00:00 GMTresearchagentsai-nativehttps://interlusion.dev/en/blog/flux-vs-argocd-which-gitops-tool-is-built-for-agents/https://interlusion.dev/en/blog/flux-vs-argocd-which-gitops-tool-is-built-for-agents/Both are CNCF Graduated, and both do GitOps well. Their architectures disagree about who the primary operator is. That matters once agents enter the loop.Mon, 16 Mar 2026 00:00:00 GMTengineeringtoolsai-nativehttps://interlusion.dev/en/blog/the-renaissance-of-angular/https://interlusion.dev/en/blog/the-renaissance-of-angular/Angular spent years as the framework people loved to leave. Then it quietly shed everything that made it heavy, added everything that makes it modern, and became the frontend framework best suited for the age of AI.Mon, 16 Mar 2026 00:00:00 GMTengineeringopinionuihttps://interlusion.dev/en/blog/arc42-the-architecture-doc-template-worth-keeping/https://interlusion.dev/en/blog/arc42-the-architecture-doc-template-worth-keeping/Most architecture documentation is either absent or abandoned. Arc42 gives it structure that humans can navigate and AI agents can maintain. Here is how we use it and why it stuck.Sat, 14 Mar 2026 00:00:00 GMTengineeringtoolsai-nativehttps://interlusion.dev/en/blog/docs-are-fun-again/https://interlusion.dev/en/blog/docs-are-fun-again/Internal docs used to rot the moment they were written. Now AI writes them, maintains them, and uses them as context. The oldest chore pays for itself.Fri, 13 Mar 2026 00:00:00 GMTai-nativeengineeringopinionhttps://interlusion.dev/en/blog/calling-llms-stochastic-parrots-is-not-analysis/https://interlusion.dev/en/blog/calling-llms-stochastic-parrots-is-not-analysis/The parrot metaphor confuses mechanism with capability. Statistical process describes how a system works, not what it can do. The debate deserves better than a five-year-old animal comparison.Thu, 12 Mar 2026 00:00:00 GMTopinionphilosophyhttps://interlusion.dev/en/blog/context7-and-the-infrastructure-layer-your-ai-agents-are-missing/https://interlusion.dev/en/blog/context7-and-the-infrastructure-layer-your-ai-agents-are-missing/Context7 fixes the context pollution that makes agents hallucinate APIs. A supply-chain flaw raises the deeper question: who fills the context window?Thu, 12 Mar 2026 00:00:00 GMTagentsengineeringtoolshttps://interlusion.dev/en/blog/your-ai-agents-are-failing-because-you-never-onboarded-them/https://interlusion.dev/en/blog/your-ai-agents-are-failing-because-you-never-onboarded-them/Prompt engineering was the warm-up act. Context engineering decides what reaches the model, when, and how. That is what makes agents drift or compound.Wed, 11 Mar 2026 00:00:00 GMTagentsengineeringopinionhttps://interlusion.dev/en/blog/orm-sucks-what-are-your-options/https://interlusion.dev/en/blog/orm-sucks-what-are-your-options/The object-relational impedance mismatch is real, ORM query languages add complexity instead of removing it, and AI agents write better SQL than ORM code. Here are the alternatives.Mon, 09 Mar 2026 00:00:00 GMTengineeringopinionhttps://interlusion.dev/en/blog/the-best-language-for-agents-is-not-your-favourite/https://interlusion.dev/en/blog/the-best-language-for-agents-is-not-your-favourite/When agents write and maintain API backends, the criteria shift: predictability, fast feedback and low ambiguity beat elegance. That points to Go.Fri, 06 Mar 2026 00:00:00 GMTagentsengineeringopiniongohttps://interlusion.dev/en/blog/what-ai-native-actually-means/https://interlusion.dev/en/blog/what-ai-native-actually-means/AI-native is not a marketing label. It describes a fundamentally different way to build software, run operations, and ship products. Here is what it looks like in practice.Thu, 05 Mar 2026 00:00:00 GMTai-nativeopinion