Day 31: Three Talks in Ten Days
v0.2.0 shipped yesterday. Tonight I stood on stage for the third time in ten days.
v0.2.0: The Enterprise Release
This is the release that makes Pinchy real for companies. Not a toy. Not a prototype. Something you can actually put in front of a team.
The headline feature is Role-Based Access Control with Groups. You can now define who has access to which agents and what they're allowed to do. An HR team that can only talk to the HR agent. A finance group that sees the accounting agent but not the engineering one. Granular, per-user, per-agent permissions. This was the #1 ask from every single pilot customer conversation.
Second: the Enterprise License Key System. JWT-based, fully offline-capable. No phone-home, no license server, no "oops, the internet is down and now nobody can work." Generate a key, paste it in, enterprise features activate. Key expires? Graceful degradation. Nothing locks out. Renew and you're back.
Third: the Audit Trail got a serious upgrade. Typed payloads with an ESLint guardrail that enforces consistency at build time. Plus we now sanitize sensitive data in audit logs automatically. Because "we log everything" means nothing if API keys end up in the logs.
And then the boring-but-essential stuff: 19 dependency updates including a Next.js security patch, Vite 7→8 migration, and a CI link checker so broken docs links get caught before they ship.
We followed up the same day with v0.2.1, fixing a Docker permissions issue and some documentation inaccuracies. Small things, but the kind of things that matter when someone is evaluating your product for the first time.
The Freelancer Meetup
The Vienna Freelancer Meetup is different from dev meetups. The audience isn't just developers. There are designers, SEO consultants, copywriters, project managers. People who build businesses around their skills, not around a specific technology.
That changes how you talk about a product. Nobody cares about your RBAC implementation or your Docker isolation. They care about the story. Why did you start? What did you learn? What would you do differently?
The talk was called "Build in Public: A Freelancer's Playbook for Launching an AI Product in 30 Days." Fifteen minutes of honest retrospective. What worked, what didn't, where the demo calls came from (spoiler: not Austria).
The Star Challenge: Round 2
At the Vibe Coding meetup last Tuesday, we put a QR code on screen and challenged the audience to star the repo. Result: +20 stars in one evening.
Tonight we ran the same challenge. "Can you beat the Vibe Coding meetup?" Final score: +10 stars. The Vibe Coders win this round. But 125 total stars from a project that's 31 days old? I'm not complaining.
What's interesting is the difference. A room full of developers who live on GitHub? They star a repo without thinking. A mixed freelancer audience? Half of them probably haven't logged into GitHub in months. The ten who did star it are genuinely curious about what we're building.
The Conversations That Matter
Three talks in ten days taught me something. The slides are the least important part. The conversations after the talk are where things happen.
Two people at tonight's meetup want to contribute to Pinchy. Not as users, as collaborators. One wants to build an integration for their workflow. Another sees a use case in their consulting practice.
That pattern keeps repeating. Every time I present Pinchy, people don't just say "interesting." They say "I need this for..." and then describe a use case I hadn't considered.
Meanwhile, at GTC
While I was preparing slides, Jensen Huang stood on stage at GTC and said something remarkable: "Every company needs an OpenClaw strategy."
Nvidia announced NemoClaw, an enterprise security layer for OpenClaw. Sandboxing, policy enforcement, controlled execution. Sound familiar? That's the exact problem space Pinchy operates in.
Some founders would panic. A multi-trillion-dollar company entering your market? But NemoClaw is infrastructure. It's a runtime, a sandbox, a policy engine. No UI. No user management. No chat interface. It's the foundation, not the building.
Pinchy is the building. And now Nvidia just told the world that the foundation matters. I wrote a deeper analysis of how NemoClaw and Pinchy complement each other.
Day 31
v0.2.0 is live. Three talks are done. 125 stars. A growing list of people who want to build with us. And the CEO of Nvidia just validated our market on the biggest AI stage in the world.
Not a bad Thursday.