Knowledge space · curated reference
happyin.space
A living library of 785+ curated articles across 26 domains — AI, engineering, narrative, strategy. Agent-ready, human-legible, no login, no dark patterns.
I’m Anastasiia — an independent builder working at the intersection of AI, knowledge architecture and developer tools. I curate happyin.space (a 785-article reference space for agents and humans) and ship open-source infrastructure for teams that use Claude Code seriously. Everything I make is free, file-based, and designed to outlive the session that produced it.
Everything I ship is part of the same conviction: knowledge and tooling should live as plain files, under your control, legible to the humans and agents who read them. Pick one to dive into.
A living library of 785+ curated articles across 26 domains — AI, engineering, narrative, strategy. Agent-ready, human-legible, no login, no dark patterns.
Multi-session collaboration for Claude Code. Atomic locks, handoffs, memory graph, messaging, voice I/O. File-based, zero core dependencies, 190+ tests.
17 architectural principles, 5 safety hooks, 16 skills and starter templates for Claude Code agents. Drop it into your project — your agent inherits battle-tested discipline.
If you’re considering working with me, or picking up one of my tools, these three commitments are the through-line.
Every tool I ship lives as plain text in your repo. No accounts, no dashboards, no vendor lock-in. You can delete my package and still read everything it wrote, forever.
happyin.space has 785 entries because each one earned its place. Same for code: I’d rather ship 100 useful lines than 10,000 defensive ones. Edit hard; ship clean.
Everything I write has to work in two reading modes — a human on Monday morning and a language model doing retrieval at 3 AM. That discipline forces clarity.
If you’re a team adopting Claude Code at scale, a researcher working on agent infrastructure, or a founder rethinking how knowledge lives in an AI-first company — I’d love to talk.