Federico Negro · 2026-04 · 4 min read
Architecture Studio's First Month: 100+ Stars on GitHub
Architecture Studio hit 100 GitHub stars in its first month. Here's what the numbers say about how architects are adopting AI tools — and what surprised us.
The numbers
We open-sourced Architecture Studio on March 2, 2026. Five weeks later:
- 102 GitHub stars and 30 forks
- 223 unique clones — people downloading and installing the repo
- 1,452 unique visitors to the repository (4,743 total views)
- 647 reactions, 50 comments, and 40 reposts on the LinkedIn announcement, with 38,108 impressions
For a niche open-source project targeting architects and interior designers — not developers — these numbers surprised us.
What the data tells us
The Grasshopper analogy landed
The announcement post framed Claude Code skills as the next Grasshopper — a way for deep experts to package knowledge and logic in shareable, reusable form. That framing resonated more than anything technical we could have said. One commenter called it “the most precise framing I’ve come across for what skills in Claude Code actually mean for practice.” The analogy gave architects a mental model they already understood.
LinkedIn outperformed every developer channel
We didn’t launch on Product Hunt, Hacker News, or any developer community. The announcement went on LinkedIn — a platform architects actually use. 38,000 impressions and 647 reactions from a single post. The comments were full of people tagging colleagues, which drove organic reach. The lesson: go where your users already are, not where tech products are expected to launch.
Stars outpace forks 3:1
102 stars vs. 30 forks. In typical developer-tool repos, the ratio is closer to 5:1 or higher. A 3:1 ratio suggests people aren’t just bookmarking — they’re forking to customize. That’s exactly the behavior we designed for. Every skill is an independent markdown file. You fork, you edit, you have your own version.
What the comments told us
The transparency rule earned trust. The post explained that every number links to its public source, every calculation shows inputs and formula, every code reference links to the government-published version. That line got more traction than any feature description. Architects are trained to verify — telling them “if you can’t see the math, the tool is hiding something” spoke directly to how they work.
It inspired others to open-source. Daan van der Zwaag from Cornell’s IPL shared that seeing Architecture Studio motivated them to open-source their own skills at gaia.computer/skills, focused on the visual concept phase. That’s the outcome we hoped for — not that everyone uses our skills, but that the pattern spreads.
People want to collaborate, not just consume. The comments weren’t “cool project” and move on. They were “let’s connect,” “can we compare notes,” “tagging my team.” Ronnie Parsons teaches Claude skills in his community. Evan Levy is building an agentic geometry kernel at Cornell. Axel Wohlin is working on jurisdiction-specific rules at Arkyv. The response wasn’t adoption — it was conversation.
What’s next
We’re focused on three things:
- More skills. The repo started with 36 skills and is now at 38. We’re adding specification writing, material takeoffs, and code compliance checking.
- Canoa integration. Our product database is coming online with 20,000+ furniture symbols and full spec data. Every FF&E skill will query Canoa natively.
- Team distribution. The distributing skills to teams workflow is getting simpler — we’re working on one-click install for Claude Desktop users who never want to touch a terminal.
The repo is public and MIT-licensed. If you’re an architect, designer, or builder curious about what AI can actually do for your practice — start here.