mcrowe

They've exposed the system prompt via the network requests. I'll give them the benefit of the doubt and assume that was intentional, and not because the app was blindly vibe coded ;-).

Here's the prompt as of around release-time yesterday. You can verify in your own network requests.

https://gist.github.com/mcrowe/9da081f52e0740886d39e730852db...

For those of us who are implementing similar AI-products, it's an interesting opportunity to see what worked for someone else. While it may be vibe-coded, it also... works.

In this case, I was interested in how they thought about the design aesthetics, and the prompt led me to the specific "skill", which the agent was happy to spit out for me when asked...

The Claude Design tool exposes its system prompt via the API requests it makes. I'll give them the benefit of the doubt and assume that was intentional, and not because the app was blindly vibe coded ;-).

I've linked the prompt as of around release-time yesterday. You can verify in your own network requests.

Happy to hear your thoughts and critique!

Thanks! No personalization on taste yet. "Personally", I'm not saving items that I necessarily like more anyways, but rather ones that I want to check out later. Future work to add an "upvote" button that drives personal recommendations on taste, though!

Thanks! Great to hear!

Yeah. Definitely a good idea. I'm told the clues are a bit cryptic as they are. It would be nice to have another if you're really stuck.

Haha. It's hard to give hints without giving it away to everyone else. Feel free to pm me on twitter @mitch_crowe and I'll be happy to give a hint.

Nice! Thanks so much for the feedback @kinduff. Those are really good points.

Hi! I created an "Escape Site". It has a score of challenging puzzles designed to make you think outside the box, and think differently about the way websites work. The success rate is about 10%. How do you fare?

I'd love to hear your thoughts and feedback!

I'm creating Cora, a new programming language and environment optimized for learning. I would love to hear your thoughts and suggestions!

Like you, I spend a lot of time analyzing the programming language ecosystem. I created LanguageHealth to help me figure out which languages are growing or dying.

I hope you find it helpful!

Yes. I had a small amount of down time on launch day. Its been up consistently since then, though!

Tabeth. Thanks for alerting me to this. I have addressed the issue. Now all the domains you see suggested should be available.

nameql@mitchcrowe.com should do the trick.

Thanks. That's an interesting suggestion. Might be a good option for revenue.

I had 1/10 be inaccurate when I did that, which is still not good enough. Its definitely something that can be addressed. Not easy, but fixable. Thanks for bringing this particular query to my attention.

I always value critical feedback. Thank you.

I think you mean that you found a name that was suggested that was not actually available. Unfortunately, this does happen, but in my experience it is less than 1% of the time. The zone files for TLDs are missing some domains which are actually unavailable.

Sorry that you had a frustrating experience.

Yeah. I agree on both counts, though. I think .com is still far more important, and adding other TLDs shouldn't take away from its prominence. For the short term, making this super clear is a good way to go.

Wow! This is super helpful. Thank you so much for to ideas, genbit.

Loved your feature suggestions. Definitely some things to move to the top of my list there. I'm curious about what exactly you meant about "search domains based on an industry". Could you elaborate on that?

Great tips on improving the business model, too. This is super valuable for me. Thanks!

Thanks a lot for your thoughts!

You're right. I would recommend the same thing to others launching on PH from now on. I didn't have any system in place to engage with users after they try it out. My product was easy to get started with, but also easy to forget.

Glad my mistakes can be lessons for others ;-).

My short term goal for NameQL is to make a really great free offering: support more TLDs, smarter suggestions, and to check for related trademarks, etc. For profitability, I was hoping that affiliate links would be enough to generate a decent revenue. It doesn't seem to be very effective, though. I'm considering adding value-added features for a small price, but I haven't planned out what those will be.

This is really useful to know. You're right, I think an engaged user might use a product once a month or so. The challenge is to stay at the top of their minds the next time they look for a new name. Having a login with value-added features may help with that.

Yes! You're absolutely right. Probably my biggest mistake was not thinking about retention before launching on Product Hunt. I suspect that your experience matches that of most visitors through Product Hunt. Your suggestion about query updates may be a good one to try, and I will brainstorm other ways to make the product give more long-term value.

And yes, it is .com only for now (plans to add other TLDs soon). Thanks for the feedback that this could be clearer.

Thanks for your suggestions simantel! Writing about what did and didn't work for the Product Hunt launch to help others is a good idea. I really enjoyed Traction, but I should probably re-read it. Thanks for the reminder!

Thanks! That's kind of what I was going for. It's a crowded market, but I felt like there wasn't a great solution yet.

Thanks! You can request access to zone files from the domain authorities. I take the raw zone file and load it into a series of Bloom filters backed by Redis. There are a lot of domains to check against, and a probabilistic data structure is the only way to store it in memory, which you need for performance. This has the small caveat that the tool will accidentally think a domain is not available about 0.1% of the time. However, it shouldn't ever report domains as available when they are not. So that works quite well for a tool like this.

You're right. It is too many clicks to save a name you like. The feature would be more discoverable that way, too.

So glad you find it useful!

Thanks! I didn't know about namechk, actually. Definitely worth looking in to an integration, or adopting some of that functionality. Good idea!

Good idea. If you click on the name you can tap on "Save" to add it to your saved names list. Does that provide what you wanted?