Thariq recently wrote about HTML being a more effective output format from AI agents — his core point being that it helps humans "stay in the loop." Karpathy echoed the sentiment. The argument is simple: when agents produce output for humans, richer formats win. HTML is visual, navigable, and people actually read it. Markdown beyond 100 lines? Most don't.
I agree. But something else has been on my mind.
As tech and non-tech folks alike start using Claude Code and other AI tools to generate documents for each other — specs, proposals, analysis — I've noticed we're getting better at producing ideas and worse at understanding each other's. Everyone generates. Few truly read. Everyone wants to be understood; few pause to understand first.
"Seek first to understand, then to be understood." — Stephen Covey
I think this principle is harder to practice now than ever. AI validates whatever framing we give it. We feel understood by the machine, so we stop trying to understand the human on the other end.
The Pain Point
This tool won't solve that communication problem. But it aligns with what Thariq is saying — if we're going to pass ideas around, the least we can do is make them easy to consume.
What I kept seeing:
- People generate beautiful HTML files with Claude Code
- But then they're stuck — it's a local file
- Sharing means "here's an attachment" or "double-click to open in Chrome"
- Non-technical teammates can't easily view or access them
So I built htmldrop — an open source CLI + Claude Code skill that makes sharing a one-command operation.
Two Ways to Use It
Mode 1: CLI — for terminal users:
npm install -g htmldrop && htmldrop init # one-time (email + verify)
htmldrop push my-spec.html # → shareable link
htmldrop push secret.html --password xyz # → encrypted, password-protected
htmldrop push internal.html --noindex # → public but hidden from crawlers
Mode 2: Claude Code skill — conversational, guided flow:
Just say: "Share this HTML file"
Claude guides you: public or private? password? block crawlers? → done.
The skill is especially useful for non-technical users — no flags to remember, just plain English. It asks what you need and handles the rest.
Privacy
Password protection encrypts the entire HTML content (AES-256) before upload. Even if someone finds the URL, they see a password prompt and encrypted gibberish. No AI agent or crawler can index the content.
Open Source
Both are fully open source (MIT). No platform lock-in, no subscription, no data on someone else's server. Use it, break it, give feedback, contribute.
I wrote a longer piece about the communication angle — how AI is making us better at producing but worse at understanding each other, and where I think this needs to go next.
Top comments (0)