Connect with the Product Hunt community

It's just over a year into Pretty Prompt, and every week we learn more about what users want.

In result, every week we reinvent the company.

Here s are my favourite things we shipped last week: (quick video here)

It s been 4 hours and Oxlo.ai is ranking #1 on Product Hunt!

A huge thank you to everyone who has supported us so far, we truly appreciate it

If you re seeing this now, don t forget to check us out and drop an upvote.
Every bit of support helps us keep the momentum going!

https://www.producthunt.com/prod...

I'm curious what websites people struggle with the most when collecting data.

Is there a website you regularly copy and paste information from into Google Sheets or Excel?

It could be an e-commerce site, a company directory, a real estate listing, Yahoo Finance, LinkedIn, government portals, job boards, or something else.

What data are you trying to extract, and what's the biggest challenge with your current workflow?

Netfox 0.11.2 is out!

Now speaking 7 languages Portuguese and Dutch join in. Netfox follows your Mac's system language automatically, nothing to switch.

When we shipped v1.0.4, the most common reply was some version of the same thing: "my subscriptions aren't all in one inbox."
Personal Gmail for Spotify. Work account for the SaaS tools. An old iCloud address still getting App Store receipts from 2021. One inbox was never going to find all of it.
So v1.1.0 is about that.
**Scan up to 9 inboxes.** Connect up to 3 Gmail, 3 iCloud and 3 Yahoo accounts.Every result lands in the same subscription list no switching profiles, no separate accounts, no re-doing your setup.

Got some feedback on the UX of Bitgrain Studio from the users, it really needs some UX improvements and redesigns. Let's see where it goes : )
When was the time you realised, your product needed some redesign?

I've been gearing up to submit agentcad to cadbench which is an eval set for CAD models.

As I've been going through that, I've found some error scenarios where everything "looks" right to me and to the agent but is actually off by a small amount here or is missing a specific shape over there.

1. Georgia: A private breeder needs a platform to find trusted sellers who work for a commission. One helper at 10% commission already boosted sales. Need more people like that.

2. USA: Need a super simple AI agent that learns by watching your screen and automates routine tasks. Existing solutions are complex or expensive. Willing to pay $10 30/month.

3. Russia: A psychologist needs an advertising bot that analyzes her workload and automatically adjusts her ads. She is willing to pay for a solution and is open to discussing the details.

4. Serbia: A photographer moving to the US needs clients. Platforms require reviews, agencies are expensive. Wants a website + ads that bring bookings. Willing to pay $100 300/month.

5. We conducted research related to this problem "Managing Type 1 Diabetes Without AI Assistance." We found 23 people from different countries and spoke with each of them. They all confirmed that they experience this problem and that it is acute for them. We will try to prepare a document with all the details in the near future.

Guys, we are also currently in the process of relocating to another country due to the war in our country. This process takes a lot of time and energy, so our activity in finding problems for ProblemHunt has noticeably decreased, but we plan to resume it at the beginning of this fall. Thank you for staying with us.

What Makes a Great Safari Extension? To answer that question, I hand-picked seven of the very best Safari add-ons for:

• Privacy

• Shopping & Deals

• Reading & Social

• Productivity

• Customization

• Search

• Passwords

What I found is that not all extensions are created equal.

The one where Skim finally admits your mailbox didn't begin 500 emails ago. Turns out there was mail before that. Who knew.

New

• Your whole mail history, actually downloaded. Until now the first sync grabbed the newest window (500 in the inbox, 200 elsewhere) and then only ever looked forward. Everything older sat on the server, invisible to search and to the assistant, forever. Skim now quietly walks each folder backwards in the background a bounded slice per pass, resuming exactly where it left off across restarts until it hits message number one. New mail never waits in line behind it.

• Install it the nerdy way. This repo now doubles as a Scoop bucket, updated straight from the release pipeline. (A WinGet manifest is wired up too, but it's still waiting in line at microsoft/winget-pkgs. Patience. )

Fixed

I remade the video to show the full BrickSolvr experience - from choosing or uploading a 3D model to customizing the bricks, exploring the result, building it layer by layer, and downloading the complete parts list.

The video is now embedded in the Getting Started guide, too.

Hey everyone,

We just launched Katalyst for sales teams on Salesforce.

Simple idea: reps don't hate selling, they hate the hours they lose every week feeding the CRM. Logging calls, fixing stages and close dates, writing next steps, reconstructing what happened on a deal from three weeks ago. The CRM they "quietly hate."

Most tools just made data entry slightly less painful. It's still the rep doing the work.

A few months ago, we exited Contral.

Contral was the first real startup Samagra and I built together. We were 18, figuring everything out as we went, and somehow reached 1500+ users and became #1 Product of the Week twice.

More importantly, Contral taught us how to ship quickly, listen to users, launch publicly, and keep building when nothing is certain.

While building it, one problem kept hurting us: consistently finding and reaching the right customers.

Yesterday, we won Product of the Day!

More than the ranking, we got to share Upstream with hundreds of founders, marketers, and product builders. That meant everything.

Two months ago, we launched Ota on Product Hunt for the first time.

It was an amazing experience. We met developers, founders, and open source maintainers who helped us see the problem more clearly.
Since then, we've shipped dozens of improvements, refined the product, improved onboarding, and listened closely to feedback from real repositories.

So we're coming back this Friday.

If you've ever joined a repo and wondered "what am I actually supposed to run?" or "why does this work locally but fail elsewhere?", that's the problem we're tackling.

Hey builders

We just added Obsidian export to Browser Notes, and wanted to share what s new + the fastest way to use it.

Built customer support AIs? Then you know the problem. You keep adding rules, paste in a policy here, teach it something there, and after a while you've got 40+ rules stacked up and rule 12 is quietly fighting rule 38. Prompt gets bloated. AI starts making stuff up.

v1.2 is our fix for that.

It learns from your team on its own
Agent replies to a customer, AI turns that into a clean rule in the background. Checks it against everything else first so it doesn't save a duplicate or a contradiction.

Just saw reports that SpaceX acquired Cursor in an all-stock deal valued at $60 billion.

This is one of the most insane startup outcomes we've ever seen:

• Cursor was founded only ~4 years ago.

• The reported price tag would make it one of the largest startup acquisitions in history.

• It went from "just a VS Code wrapper" jokes to becoming the default AI coding tool for a huge number of developers.

• SpaceX/xAI would instantly gain access to one of the most valuable datasets in AI: how millions of developers actually write, edit, debug, and ship code.

Someone commented on our launch: "the gap between 'wrote it down' and 'actually did it' is where every note I've ever taken goes to die."

That's the whole reason Buddy exists. I have a graveyard of half-thoughts in my note app that never became anything.

So, genuinely curious: where do your notes go to die? Notion you never reopen? A Slack message to yourself? Sticky notes? Tell me your worst one

Most AI video tools I've seen are built around one thing: generating new footage. But the problem most creators and editors I talk to have isn't a lack of footage. It's figuring out what to do with the footage they already have.

Curious what the community thinks: If you had an AI editing assistant, what's the one part of editing you'd want it to handle for you?

Nine years of working closely with insights teams across CPG, BFSI, and tech has given us a consistent observation: the transcript is where researchers start, but the recording is where they go when the transcript isn't enough.
They go back to hear how someone said something. To check whether the pause before an answer was genuine uncertainty or just thinking time. To see whether the person's expression changed when they described a feature they claimed to like. To catch the moment when engagement visibly dropped when they stopped leaning forward, when their eyes moved away from the screen even while their words stayed polite and positive.
That gap between the transcript and the recording is where a significant amount of research insight lives. And it's also where the most time gets consumed in manual analysis. Researchers who run 20 or 30 interviews know what it's like to sit through hours of footage looking for the three or four moments that actually matter.
This observation that the transcript captures what was said but not how or with what underlying feeling is fundamentally what drove us to build the emotion and behavioural layer into Mira. Not to replace researcher interpretation, but to give researchers a way to find the right moments faster.
I'm genuinely curious: what's the signal you find yourself wishing you had after you've finished going through a set of interviews? What consistently drives you back to the recording that the transcript alone doesn't give you? And have you found any workarounds annotation practices, tagging systems, collaborative review setups that help you capture more of what the transcript loses?

But you can only spend it on ONE thing.

What are you choosing?

This should be interesting.
Let's goo

I will go first.

"Do you do backlinks?"

Every week. Sometimes every day. From prospects. From agencies. From people who read one SEO blog post from 2018.

Backlinks are not the problem. Backlinks are a solution to a problem that has changed.

Every AI product I've used starts the same way: an empty chat box and a blinking cursor. The tool has all the capability in the world, but it sits there waiting for me to figure out what to ask.

That's a design failure we've all just accepted. The blank page problem killed writing apps for years, and now we've rebuilt it into every AI interface.

Something I built led to a design decision I want to get feedback on.

LineageLens is a free VS Code extension that captures every AI code insertion and scores it for risk on a 0 100 scale. Works with Cursor, Copilot, ClaudeCode, Gemini CLI. Zero config on install just start using your AI tools and your insertions start showing up in the sidebar.

The scoring is deterministic rules: +28 for credential-like material, +24 for eval/exec patterns, +22 for subprocess calls, +14 for landing in an auth or payments file, and so on. Fully traceable. No ML, no black box.

The design decision that surprised me: missing prompt capture when the extension records a file insertion but has no record of what was asked adds +24 to the risk score. Same weight as detecting an eval() call.