error404x

model I used for preset generation was gpt-5-mini. I’ve changed it to another model. Hopefully, it will follow the instructions better.

Thanks guys for the feedback. I’ve changed the preset generation model from GPT-5 Mini to GLM-5. The quality has improved a little, but the model is slightly slower

I’ll fix it. Sorry about the issue.

Yep, it’s just a vibecoded project, and the model used for the preset isn’t great since I’m using gpt-5-mini. I’ll change the model. It’s just a fun weekend project only

will fix it :)

Thanks for trying it out. I think I need to change the model that generates the preset. It’s currently using gpt-5-mini.

This was just a fun experiment where we played around with the figma api, but their api has been rate limited since nov

Yes, we currently use FSL-1.1-MIT, but we’re open to changing it to AGPLv3 or another open-source license.

I built a small url shortener and also experimented with a map‑based mood tracker. what stood out to me is how quickly I could go from a prompt to a working frontend + backend without boilerplate.

For me, the most useful next nodes would be: 1) auth 2) stripe 3) file upload 4) convex action nodes (for more complex workflows)

I've been playing around with vibeflow for a while, it's impressive how fast you can go from a prompt to a working full stack app. The visual workflow editor is a game changer.

Good, no problem. I just shared what I saw on Hugging Face.

Hey hn,

I created this because I love exploring how simple changes can completely transform something, especially when it comes to numbers. It's a fun way to challenge your thinking by mutating digits step by step to reach a target sequence. Hope you enjoy it!

Hey HN!

I made a small side project: https://recipetolist.com. It's a website that takes a recipe URL and converts it into a clean, structured grocery list. Just a fun project

give it a try!

I still prefer using search engines, but if I don't get an actual answer, I use LLMs with internet search. It is sometimes helpful for me, though most of the time, I get similar results as a normal search engine.

Some AI search companies [0] are even planning to add ads to their results, possibly on their free plans, which could make it harder for ad blockers to filter them out.

[0] https://techcrunch.com/2024/11/12/perplexity-brings-ads-to-i...

I tried crossing my eyes, but it’s not working for me; I keep seeing things blurry. Maybe I’m doing it wrong. However, I solved the first two puzzles. For the last one, I just guessed randomly. My guess wasn’t exactly correct, but it was close, just a little distance away.

I used Koa a few years ago. It's good, but meh it’s not really the 'Next generation web framework for Node.js.' I still use Express and sometimes Fastify these days.

This is so amazing, thanks for sharing!

I usually create an API key like this: `sk_${randomUUID().replace(/-/g, '')}`.

Thanks for the feedback! I'll work on improving it. :)

Thanks! I will add captchas too.

Sorry, I just added '1999' as a meme for retro design. Apologies for any confusion caused.

Sorry, CAPTCHA caused issues. My other project got a bot attack, so I have added CAPTCHA to all my projects. I didn't realize captcha was causing problems to users

Omg, I didn't check HN for a few days. Sorry, Turnstile was causing issues for everyone, so I have removed it for now. This is just a fun project :)

Thanks for pointing that out! It was a mistake to include Vatican City in the list since it’s a UN non-member observer state. I see how that made the question and answers a bit redundant—I'll make sure to fix it!

Hey, HN, this is a small side project of mine. Dialoqbase-lite is a simple ChatGPT-style web UI and side panel chrome extension. Please try it out and provide valuable feedback. Thanks!

Extension: https://chromewebstore.google.com/detail/dialoqbase-lite-a-s...

Update: Since my ElevenLabs credit ran out, I'm now using the SpeechT5_TTS model.

The whisper I'm using is self-hosted on the same VPS, but yeah, I will try llava vision. Thanks

Note: Currently, the application does not load PDFs or YouTube videos.

Also Video processing may take a few minutes because of my VPS. I'm welcoming feedback. Thanks :)

This is just a fun project; don't take it seriously.