Hi Product Hunt!

After spending months building and experimenting with AI agents, we kept running into the same problem.

The models were getting smarter.

The infrastructure wasn't.

AI tools generate ideas faster than anyone can sort them. I kept ending up with Slack threads, half-finished docs, and twelve tabs of "things to look into" - and no clear picture of what was actually shipping.

So I built agency-os. It turns Notion into a single board for working with AI:

• You talk through ideas with the agent. It pushes back, asks clarifying questions, and scopes the work into tasks and subtasks on the row.

• You approve before anything runs. Nothing dispatches autonomously - every execution is operator-gated.

• Agents pick up approved rows, work in parallel respecting dependencies, and close them with result links.

The board is the source of truth. The agent is just an executor.

Hey Product Hunt community

I'm a 19-year-old national athlete from India. I started trading 2 years ago. Blew up 3 accounts. Lost 4 lakhs.

Not because of bad strategy. Because I couldn't follow my own rules.

Moving stops. Adding to losers. Revenge trading. 23 violations in one month. $3,200 lost.

Hey Product Hunt,
Solo dev here. This started with a simple problem: I needed SEO tools and couldn't justify $99 $399/month for them. So I started building my own.
Somewhere around tool #20, I noticed something that bothered me about the tools I was replacing. A lot of them show you numbers search volume, difficulty scores, "AI visibility" without ever telling you where those numbers come from. Sometimes it's real API data. Sometimes it's an AI estimate wearing a confident font. You can't tell the difference, and I think that's the point.
So I made one rule for myself: every tool has to wear a badge saying what it actually is.
Live real data from real APIs
Live + AI real data, AI analysis on top
Generation AI wrote this, treat it like a draft
Analysis AI opinion, not measurement
If a tool can't get real data (say, you haven't added an API key), it refuses to fake it. It literally returns an error saying "I won't estimate this and pretend it's live." I lost count of how many times I was tempted to soften that. I didn't.
115+ tools later, it's a full platform: keyword research, content generation, technical audits, an AI site auditor, and the thing I'm most excited about a monitor that tracks whether ChatGPT and Perplexity actually mention your brand, with public scorecards anyone can check.
Free tier is 5 generations a day, no card. If you bring your own API keys, it's basically unlimited you pay OpenAI/Anthropic directly instead of paying me a markup.
It's seofragments.com if you want to poke at it.
Two honest questions for this community:
1. Do honesty badges actually matter to you, or do users just want the number and don't care where it came from?
2. If you've built something with a "principled" constraint that made your life harder did it pay off, or did nobody notice?
I'll answer everything in the comments.

I've been building Fog, a notes app that tries to solve a problem I kept running into: notes apps either make you organize everything manually (folders, tags) or don't organize at all.

Fog uses Apple's on-device Foundation Models framework (the same one powering Apple Intelligence) to read your notes and group related ones into "Clouds" automatically no manual tagging, and nothing ever leaves your device since it's all on-device inference.

The interesting technical challenge was the clustering logic. I ended up using a union-find (disjoint set) structure combined with NaturalLanguage embeddings to figure out which notes are actually related, then merge them into clusters that update as you add more notes. Getting the similarity threshold right so Clouds don't over-merge unrelated notes or fragment obviously-related ones took a lot of tuning.
Demo video (56 sec, shows the auto grouping)
Curious if anyone else here has worked with Apple's on-device Foundation Models yet how are you handling similarity/clustering tasks, and did you run into similar tuning headaches with thresholds?

Been running a phone and electronics repair shop for a few years and tried a bunch of different tools to manage tickets, inventory, and customer records. Most were either too complex or just not built with repair shops in mind.

Stumbled across Manage Repairs (managerepairs.com) a while back and it's genuinely been the best fit. It handles repair tickets, POS, inventory, and customer history all in one place. Clean interface, nothing bloated, and it just works the way a repair shop actually runs.

Not here to spam just genuinely think it's one of those tools that doesn't get talked about enough. If you run a repair shop or know someone who does, worth checking out.

Planning birthdays, house parties, office celebrations, or festive events is still chaotic.

Too many calls. Random vendors. No transparency. No easy comparison.

So we built Tendr.

Discover trusted vendors
Compare decor, catering, DJs & photographers
Wishlist vendors before planning
Plan events faster with a smooth workflow
Designed for private celebrations & office events

We ve already onboarded 100+ vendors across Delhi NCR and validated demand through real orders.

Would love your feedback and support

Hey PH Community

oday we launched CATIA. It started as an internal frustration, and apparently we're not the only ones who had it

Every time we needed qualitative insights, the bottleneck wasn't the research.

It was everything around it.

Hey PH, I built a small app that collects and decodes Chinglish signs. It s a fun way for Chinese learners to understand common translation mistakes. Would love your feedback.

Here s the product page: https://play.google.com/store/ap...

[ ]

Education is evolving.
Why just learn... when you can play, earn, and automate?
Welcome to Rapid HQ AI
where we transform expertise into monetization using gamified technology!
Forget boring lectures. Our gameplay feels like your favorite chat-bot, but with a twist.
Every interaction is powered by hyper-personalization.
Our A-I learns your goals, adapts to your skill level, and generates a custom roadmap just for you.