Yansu, from Isoform, is a proactive AI app builder. Instead of waiting for a prompt, it observes how you work — via desktop screenshots and messaging apps — learns your patterns, and builds custom applications, workflow improvements, bug fixes, or insight dashboards on its own.
## The proactive angle
This is the same “proactivity” frontier that recent agent benchmarks show frontier models still struggle with. Yansu’s bet is that watching real work is the shortcut: don’t ask the user to specify what they need, infer it from what they actually do. It pairs with a companion skill that crystallizes the learned context and delivers it to whichever agent you’re using — Claude, GPT, Gemini — so you don’t re-introduce yourself each time.
## Privacy and models
Free on Mac, Windows, and Linux. Multi-model — each step of app generation uses whichever model performs best. Memory, conversations, and generated apps stay local; nothing leaves the device without explicit permission.
## Why it matters
“Proactive” is the most-claimed, least-delivered word in AI agents right now. Yansu’s observe-then-build approach is a concrete attempt — and the local-first, multi-model design sidesteps the two things that kill adoption of always-watching tools: privacy fear and single-vendor lock-in.

Leave a comment