Absolutely, it's also cool to see that so much is happening here right now. Sometimes you need a bit of a network to get something like this off the ground. Building infrastructure is never easy.
The idea was simple: keep AI prompts, intents, and conversations alongside your code and commits — basically treating AI interaction as first-class development artifacts. everything just plain Git
We struggled to get momentum. Things happen.
Now, less than 24 hours ago, Thomas announced Entire.io:
“Entire CLI hooks into your git workflow to capture AI agent sessions on every push. Sessions are indexed alongside commits, a searchable record of how code was written.”
That’s… very, very close to what we tried to build.
Honestly, I love the vision and think this will matter a lot in the AI age. It’s validating to see someone with that reach betting big on it.
Best of luck to Thomas and the team behind Entire.
It’s surprising how fast you can design real interfaces with Pencil.dev already. The workflow feels closer to writing code than pushing pixels, quick iterations, clear diffs, and no heavy UI getting in the way.
What really stands out is that everything lives in Git. Designs sit next to the code they relate to, with versioning, history, and collaboration handled by tools developers already use. This avoids a lot of the friction Figma had for years, where design history, branching, and reviews were either missing or awkwardly bolted on later.
If Sketch -> Figma was about moving design to the browser, Pencil feels like the next step: treating design as a first-class, versioned artifact in the developer workflow.
Curious how designers and engineers here think about Git-based design workflows, and where this approach might fall short compared to Figma.
Perhaps I should add something here. It always depends on the task. Claude with SUDO access doesn't seem right to me either, but I wouldn't run that anywhere else either.
If your system were under version control, so that Claude could do whatever it wanted on its own branch, so to speak, would it still be such a big problem? Because you could just roll back if it really did cause problems, couldn't you?
On expanding beyond Claude: there’s no concrete plan right now since we built this around Claude, but we’re very open to it. If you have a preferred CLI (e.g., Codex, OpenCode, or something else), feel free to open an issue in the repo. Or just describe your use case here and I will do it :)
Regarding branches: the tool does not pollute your working branch. Each session lives on a separate “session” branch that contains all prompts and operations. Your normal working branch stays clean.
When you end a session, you’re prompted to either:
merge the code changes into your working branch, or
discard them.
If the video or README didn’t make this clear enough, I’d appreciate the feedback I’ll update the docs accordingly.
A current concern of mine, also when using Claude Cowork, is that I don’t want files to be modified if I can’t control their versions/ if they aren’t versioned at all.
And even though it doesn’t actually get full sudo access, giving an LLM permission to edit files without being able to track exactly what it’s doing still feels risky.