Connect with the Product Hunt community

Searched the sub for agent-driven dev workflows and came up pretty empty. I mean those setups where autonomous agents write code, test it, and basically deploy small apps with minimal hand-holding.

I run a few small automated tools myself, so I'd love to swap notes with people who are actually building this way.

Is anyone else playing with this? What frameworks or orchestration setups are you running for your coding agents? More importantly, how do you keep them from losing context or breaking things when the codebase gets bigger? What decisions are you letting the agent make completely on its own without reviewing first?

I m a solo founder building with AI tools. This week I found out the emoji reactions in my app had never been saved for anyone.

The cause was simple. The code wrote to a column called reaction, but the live database column had a different name. The app still showed the emoji as highlighted, and nothing was stored. The count of saved reactions was zero, and no error showed up anywhere.

The fix was small: a migration that renames the column only if the old name exists, and a check so the app now says not saved when no row actually updates.

What I still don t have is a habit that stops this from happening again. If you build on Supabase or any hosted database, how do you keep what s live and what s in your code in sync? Do you check it before every release, or just wait for a bug?

I build a trading bot for Polymarket, so Brazil's runoff on Sunday was an interesting one to watch. The week before the vote, the market priced Fl vio Bolsonaro at 77% while Datafolha and MDA had the race basically tied. On election night, after 8:30 p.m. ET, Kalshi still had Lula at 21%. He won 51.8 to 48.2.

The next morning B3 hit its circuit breaker at the open for the first time since 2020. The Ibovespa closed down 13.9%, the dollar jumped 8% against the real, and EWZ lost 22% in New York. Stocks had been priced for a Bolsonaro win too, and the market moved all at once when that turned out wrong.

Spent a full week forcing daily updates on social media because everyone says you have to do it for traction.

Ended up totally exhausted and only got 2 signups. Felt way better to just close the laptop and fix code in silence without posting about it.

How do you guys keep sharing updates without feeling like you're just performing for an empty room?

I'm preparing to launch my first product on Product Hunt next Monday, and I've been going down the rabbit hole of launch strategies lately

Some people say to launch right at 12:01 AM Pacific. Others say the day itself matters less than having a good product, a clear message, and people who genuinely care about what you're building.

As a solo maker, I'm trying to figure out where to focus my energy instead of overthinking every little detail.

For those who've launched before:

Building everything your loudest users ask for is a classic trap. We went down that exact same path early on and ended up with a bloated app that confused everyone else. Power users want niche workflows that might not fit your actual target audience at all.

We started telling users "no" or pointing them to workarounds, and our retention actually went up because the app stayed simple.

Do you separate feedback from paying enterprise clients versus regular users, or treat all requests equally?

A user told me my product was too cheap. Not a complaint about quality. They said the price made them doubt whether it was serious.

I had spent years worrying that I was too expensive. It never occurred to me that some people read a low price as a warning sign. Cheap can mean accessible, and cheap can also mean fragile.

I check my competitor's product every morning. Not productively, not with a plan. Just a quick scroll to see what they are doing.

It is the worst ten minutes of my day and I keep doing it.

I m building an app solo, and users can share their results as a link. The page has to open for people who aren t logged in, but the data belongs to a private account.

What I did: the link uses a random code instead of the real ID, and the page only gets the few fields meant to be public. Then I tested it in a private window, with a real code and a fake one.

It works, but I m not sure it s enough. If you ve built something like this, what do you wish you d checked earlier?

Been on this platform a month now, and genuinely think the people here are sharper than what i've found anywhere else.
Curious if that's actually held true for others when it comes to hiring.
Have you hired someone you found on ph, or been hired yourself through here.
Also curious what your actual go-to platform is for finding people, ph, linkedin, twitter or somewhere else entirely.

We spent three full months building a complex permissions dashboard because a few noisy users kept asking for it.

Turns out it just cluttered our settings and scared new signups away. We took a deep breath and deleted the whole thing back to simple roles.

Support tickets dropped instantly and signup usage shot up. Funniest part? the people who begged for it barely even used it.

How do you decide when a feature needs to be thrown away rather than fixed?

Saw that new app Tab raised at a huge valuation for a bot that lives entirely inside iMessage and WhatsApp.

It actually books stuff and spends money for you instead of just giving instructions.

Not having to download a new app is great, but texting feels super sketchy when real money is involved.

How can text-based tools give you clear receipts so you know they didn't mess up your order?

When evaluating new tools, I ve started ignoring flashy marketing pages and heading straight for the product's changelog.

Seeing steady weekly fixes and listening to real feedback tells me way more about the team's momentum than a row of testimonials.

Framing every update around a user headache like "fixed slowdowns on big exports" instead of vague tech speak makes a massive difference in showing care.

Has publishing a public release history helped your conversion rates, or do you find most visitors ignore it?

For the illustrative values 10, 11, 12, 13, 54, the mean is 20 and the median is 12. A single headline number hides that distinction.

Would you show a small distribution chart next to the summary, or expose the extra detail on demand? What design has worked for readers who are unfamiliar with statistics?

Hi everyone,

I m looking for tools that can take a single image of a person or cartoon character and make them dance in sync with an audio track, so the movement follows the rhythm and beats of the sound.

I ve already tried open-source models, Wan2.2 and LTX. The results weren t good enough. The motion was off-beat or barely there, and the character s identity drifted over time.

What I need:

I hated meetings. Still do. But I noticed the problem was not the meeting itself. It was the scheduling, the setup, the thirty minutes of context before anything useful happened.

So I tried something. Instead of booking time, I sent a voice note. Just me, talking through the thing I needed to say, for two minutes.

Fixing slide alignment and spacing takes up way more time than actually writing the content.

You spend an hour tweaking margins and then someone asks to translate the whole thing into another language, which completely breaks every single layout.

I started using strict layout grids to speed things up, but multi language formatting is still a nightmare.

Do you rely on standard slide templates to keep things consistent or does every deck get a custom design?

I spent two years trying to explain what my product does. Different copy, different angles, different frameworks. Nothing landed the way I wanted.

Then a user described it to me in passing. One sentence, in their own words, using language I would never have chosen.

I had a bug that reproduced about once a week and never when I was looking for it. My tests passed. My logs were clean. Then I would get an email from a user describing something that should have been impossible.

I could not see it for a month. And then I caught it because I happened to leave a terminal open overnight and came back to a strange line of output.

I learned more from a failed launch than I have from every book I have read on the subject.

The books made sense. The principles were clean and the examples were instructive. And none of it stuck until I had felt the thing go wrong in my own hands.

Last month was slow. Nothing was working the way I wanted and every day felt like pushing something uphill.

Then someone sent a one line email saying the tool had saved them an afternoon. Nothing dramatic. No details.

I automated a task last month. Replaced about two hours of weekly work with a script that runs in seconds.

I felt efficient for about a week.

Someone asked me a technical question last month. I gave them an answer that was accurate, well reasoned, and did not help them at all.

I could tell because they thanked me politely and asked someone else.

I added a dependency last month because it solved a problem in twenty minutes. Saved me a day, at least.

A week later I needed to change one small behavior in that dependency. Not a big change. A small one. And the library was not built for it. So I forked it, and now I maintain a fork, and the fork is slowly drifting away from the original.

Makers: when you change your UI, what goes stale first: your demo video, your help docs, or your onboarding emails?

For us, keeping tutorial videos aligned with the latest UI is an ongoing challenge. Even a small interface change can make an older recording confusing when it shows a button or workflow that users can no longer find.