I Built My Own RSS Reader in an Afternoon, With AI Doing the Typing

A developer built LumenAI, a local-first RSS reader for macOS with AI summaries, in roughly an hour by acting as product owner and build verifier while an AI wrote all the code. The project used a phased spec with explicit non-goals and a human build check-in after each phase, since the AI's Linux sandbox couldn't compile a Mac app, yielding 38 Swift files, 27 tests and no runtime crashes. The AI layer centers on a single SummaryProvider protocol with six implementations, including Apple Intelligence on-device, Ollama, Claude, OpenAI and Gemini, with API keys stored in the macOS Keychain.

For years my RSS setup was a compromise. I never wanted to pay for a desktop reader, so I stayed on The Old Reader: a perfectly fine service, but it felt like visiting a website, because that's what it was. What I actually wanted was simple: a native Mac app, feeds stored locally, no account, no subscription, and fast enough to triage a hundred articles from the keyboard. So I stopped wishing and built it. LumenAI is a local-first RSS reader for macOS with AI summaries, taken from an empty folder to a notarizable DMG in about an hour of wall-clock time, with me as product owner and build verifier while the AI wrote the code. My pitch to the AI was one paragraph: a local RSS reader for Mac, a premium feel, and treat it like a real engineering project, with clear phases, and ask me questions before making decisions . That last sentence mattered most. Instead of a wall of code, it interviewed me: stack, what v1 includes, what it excludes, how refresh works, what "premium" means concretely. A few rounds of multiple-choice later we had a spec. Writing down the non-goals felt ceremonial. It wasn't: every time scope tried to creep, that list ended the discussion in one line. Every choice optimised for native feel, local data, no servers. .xcodeproj . The part I'm proudest of is the AI layer: one SummaryProvider protocol with six implementations Apple Intelligence on-device, Ollama, Claude, OpenAI, Gemini, and Disabled . Swapping providers is a dropdown; keys live in the macOS Keychain. for loop. <link tags, one canonical model for every feed format. j / k /space/ s / m keyboard navigation. Every phase went the same way: the AI proposed decisions and asked questions, wrote the code and tests, then stopped, because its sandbox is Linux and can't build a Mac app. So I was the build machine: ⌘U, ⌘R, report back. That constraint turned into a feature. It forced a human checkpoint between phases instead of a firehose of unverified code. Of three failures in the whole project, my builds caught two and the AI caught one re-reading its own code. Final tally: 38 Swift files, 27 tests, zero runtime crashes encountered. Treat it like an engineering project, not a magic trick. Phases, a non-goals list, tests, and a check-in after every phase are why this took an hour instead of unravelling over a weekend. The AI typed every line; the spec, the taste and the "no, simpler" calls were the human part, and the project needed both. That's the whole idea behind what I write on Curious Bit https://curiousbit.netlify.app/ : you learn a stack by shipping something real in it, not by watching a syllabus. If you'd rather learn infrastructure-side AI the same way, I've also put together AICademy https://ragentic.netlify.app/ , built around doing the job rather than reading about it. Full write-up with the screenshot gallery: I Built My Own RSS Reader in an Afternoon https://curiousbit.netlify.app/i-built-my-own-rss-reader-in-an-afternoon-with-ai-doing-the-typing/ Originally published on Curious Bit https://curiousbit.netlify.app/i-built-my-own-rss-reader-in-an-afternoon-with-ai-doing-the-typing/ .