PH热榜 | 2026-03-10

👋 Hi Product Hunt! CY here, founder of Vozo.

I’m an ex-Googler researcher who helped build core video technology for Android, Glass, and Photos.

Visual Translate is Vozo’s 3rd launch on Product Hunt — bringing the last missing layer of video translation: the text inside videos. It builds on our previous successful PH launches around AI dubbing, lip-sync, subtitles, and translation quality.

👉 Fully translated videos — finally possible.

With Visual Translate, Vozo can now translate the text inside videos — slides, diagrams, UI labels, and callouts — while keeping the translated text fully editable.

This turns out to be surprisingly tricky: the system has to decide what to translate, what to keep, and how to recreate visuals without breaking layout, style, or animation — but we’ve finally made it work.

We’re starting with slide videos and explainer videos, where much of the information lives directly in the visuals. With this final layer solved, important videos can finally travel across languages instead of being locked inside one.

🚀 We’re opening FREE beta access today — sign up with Gmail and try Visual Translate. Let us know what videos you’d translate first.

A small backstory on how Visual Translate started.

The idea goes back to October 2025. Around that time we noticed that many great educational videos weren’t being translated well. A big reason was that a lot of key information wasn’t only in the narration, but in text inside the visuals — slides, diagrams, labels, and callouts.

When we looked at existing video translation tools, we realized that this layer was still largely unsolved.

So we decided to try building it.

Huge credit to our engineer Naro. She started experimenting with the idea back in October and built the very first prototype and pipeline herself. The demo she showed the team was still rough, but the results were already surprisingly impressive.

Naro is honestly one of those engineers who are both brilliant and delightful to work with — sharp, curious, and incredibly creative when exploring new ideas. That early experiment she built convinced us this was worth turning into a real product, and the rest of the team quickly rallied around the idea.

This could save a lot of manual After Effects work.

What happens when the translated text is longer than the original space allows?

Is visual translation a separate module or part of the main workflow?

One design choice we cared a lot about while building Visual Translate is editability.

A lot of AI tools today focus on full generation. That works well for creating something from scratch, but in many real workflows people aren’t starting from zero. They already have a finished video and just need to adapt it for another language or audience.

Instead of regenerating the video, Visual Translate separates the text layer inside the video, translates it, and rebuilds it back into the visuals while keeping everything editable. You can adjust wording, layout, or styling directly in the editor.

For us, this approach fits much better with how video localization actually happens in practice. It’s been really exciting to see teams in different industries already using it for training videos, explainers, and internal communication.

Omg, living in a foreign country you have no idea how amazing this is! I am so excited to try. Do you have Hebrew?

Congrats on the launch, The demo looks great! I’m definitely interested in trying it out.

This is perfect for educational videos where visuals carry as much meaning as the narration. Congrats on the launch!

One quick question, do you offer API?

Hey team, congrats on the launch! Super polished product with a validated real world use case. Professional demo. Excited to try it out. Wondering if you offer an open API?

Really interesting approach to video localization. The hardest part of translating video content has always been that the visuals and text are so tightly coupled. Curious how it handles text embedded in complex backgrounds or fast-moving scenes.

amazing tool i love the concept

It's cool tho. Does Vozo work better for educational videos or marketing videos?

When space is limited, how does Vozo handle it? Does it prioritize readability or literal accuracy?

Congrats on the launch! Translations seem super natural! 🎉

Great use case, good luck for the launch team!

Congrats! How does Vozo fit into a typical YouTube localization workflow?

Congrats on the launch! Just tried it and loved it.

Quick question — is there an edit history for visual translation changes? When working with our review team, we usually go through several rounds of revisions before settling on the final wording, so being able to track changes would be really helpful.

What scenarios do you think Vozo works best for today, and where does it struggle the most?

Congrats on the launch @lightfield 🎉

Can Vozo translate text that appears for only a few frames?

Congrats and good luck! Very much needed tool in our global markets!

Can Vozo translate screenshots embedded inside videos?

Does it preserve Voice and emotion? or it sounds like Netflix's international movie dubbing ? :)

How does Vozo handle very small or faint text?

Hey, congrat for a launch

Love this. The on-screen text translation is the piece most video localization tools completely skip over. Being able to translate slides and diagrams inside the video without rebuilding the visuals is a huge time saver. Curious how it handles text that's baked into animations or motion graphics?

Tried Vozo and was really impressed by the lip-sync accuracy—it’s a huge step up from generic tools! My main curiosity is around edge cases: How well does the model handle profile shots or moments of high emotion (like shouting or laughing) where mouth shapes are very dynamic? Curious how robust the "human-level" sync is in those tricky scenarios.

Cool product! This can truly help scale video to a broader audience. How long does it take to process a video in multiple languages at once?

Hey Product Hunt family 👋 We’re back!

A few months ago you helped Chronicle become #1 Product of the Month, and later #4 Product of the Year. Since then, the product has grown to 200k+ users. With that, came a huge amount of honest feedback. Hundreds of you told us where Chronicle worked, where it didn’t, and where it could be much better.
We read all of it.

Today, we're excited to launch Chronicle 2.0 to all of you!
Chronicle turns notes, prompts, or existing slide decks into stunning, on-brand presentations, generates an impressive first draft, and lets you refine slides through conversation.
There are a lot of AI presentation tools popping up right now. But most of them generate what we’ve started calling AI slide slop: generic layouts, messy structure, and decks you still have to spend hours fixing.
Our ambition is simple: to build the best storytelling tool for teams.

Here are a few things I’m most excited for you to try:

✴️ Chronicle AI: Think of it as a slide design coworker. Give it anything: notes, a prompt, or a rough idea, and it builds an impressive first draft you can refine and iterate on together.

🖼️ Custom branding & themes: Dozens of beautiful templates with real design taste, plus full brand customization so teams can stay on-brand.

📊 Charts & graphs: You can now add and customize charts directly inside Chronicle. They automatically adapt to your presentation’s theme.

🗃️ Export to PowerPoint & PDF: One of the most requested features, now live.

🪄 Hundreds of world-class templates: Created by designers from places like Apple, IDEO, McKinsey and BCG.

Special thanks to @benln for hunting us. And honestly, thank you. The feedback from this community is the reason Chronicle 2.0 exists. As a small thank you, use code PHPRO this week for 1 month of free PRO on any new account.

We’re shipping updates every single day, and will be hanging out in the comments all day. We’d genuinely love your thoughts. What should we build next?

PS. We also have a Slack community where you can interact with the team directly, get early access to new features, and shape what we build next. Would love to see you there: chr.so/slackcommunity

Proud to be part of this team! From day one, the ambition has been to build something truly great for storytelling.

What we’ve created has the potential to change how people tell stories - making it easier, more powerful, and more impactful than ever.

Huge moment for the team today. Can’t wait to see what people build with Chronicle ✨

I knew this launch would be special, we've been listening to your feedback and created a tasteful way to create presentations. Templates, image gen, themes, and new widgets are just the tip of the iceberg.

I'm so proud of this team and can't wait to hear what you all think! Would love to hear your feedback!

Seeing Chronicle launch today is really special 😇 — this team has spent a lot of time thinking about how ideas should be communicated visually.

Really proud of what the team has built here. Chronicle is the result of a lot of thoughtful design work and storytelling thinking.

Honestly one of the most fun products I’ve ever had to market. As a marketer who spends all day thinking about storytelling, working on a storytelling product is a dream 💖 Can't wait for everyone to try it!

It’s been inspiring watching the team build something that puts storytelling and design quality first.

Hey Product Hunt

Mayuresh here 👋

Just wanted to jump in and say a massive thank you to this community. The previous launch was my first ever product launch yet - seeing the love and support was really special.

In the last few months, we have seen 200K+ users doing sales, proposals, research reports, QBRs, MBRs, all hands and more with Chronicle! We are back with tons of improvements and some big upgrades.

1. We have built a whole new generation experience with AI that works like you are used to (digests files, looks up, goes back and forth to iterate with you). We rebuilt this ground up to make sure it is truly powerful.
2. We have added tons of small features that give you speed and customisability, made overall editing easier and faster (e.g. gridlines, better colours, better themes, tidy up options)
3. Export to PPT is in :) its a first cut - we will continue improving fidelity. There's more export options too.
4. We built our own charts - fun story @praveendinesh shipped this literally in 2 days.

A lot of the improvements in Chronicle 2.0 came directly from all your feedback, so please keep it coming. We read everything.

Excited to hear what you think!

Really excited to see the all new Chronicle out in the world today, this team genuinely wants to build a generational storytelling product.

Love the idea of combining AI generation with real editing control.
That’s usually where many AI tools fall short. As someone designing AI-powered products, I’m curious:
Does Chronicle generate slides within a consistent design system, or can teams plug in their own brand styles/templates?

Launch day!! Really excited for people to finally try Chronicle.

Really excited to see the all new Chronicle out in the world today, this team genuinely wants to build a generational storytelling product. Really happy and proud to be part of a super talented and passionate team of people.

The positioning around AI slop is spot on. Every AI presentation tool right now generates the same generic layouts with the same stock-photo energy. The fact that you are leaning into design quality and conversational refinement rather than just speed is a real differentiator. Congrats on the 2.0 launch and 200k users!

So awesome to get this out today. I love that we have visual AI with taste now!

I use Chronicle extensively - great product. It's replaced most other deck creation tools. It's a core part of my workflow now.

How does Chronicle’s widget-based architecture ensure that layout integrity and interactive elements remain intact when exporting to static or legacy formats like PPTX and PDF, and does it allow for the ingestion of existing corporate master slides to maintain strict brand compliance?

Super excited for this! Grateful to everyone who shared feedback since our previous launch - a lot of it directly shaped what you see today. So lucky to be a part of this team ♥️

This launch is very special for the team. Months of work behind the scenes and years of data and learning finally shaping the way Chronicle is today. Huge shoutout to our engineering team, Claire and Elliot. I am in love with the new version. Can't wait for the world to experience this. 🔥✨

I really like how Chronicle is focused on how information flows throughout the deck. Especially since a slidedeck is meant to tell a story.

Are there ways to give further context to Chronicle such as the specific target audience, tone, or even the type of narrative/story I want my slides to convey? For example, "a startup pitch deck that is meant to be playful but concise and to the point for investors"

I've used Chronicle, and I've used Gamma, but I felt that Chronicle fits my vibe much better than Gamma.

Both tools are good, but for me, Chronicle is a better fit. I really enjoy using the app, but if there's one thing I would improve, it is the customer support. I feel the customer support has a lot of scope for improvement. And a bad experience (delayed response) can leave a bitter taste.

Anyhow, wish you guys the very best! Onwards and upwards...

This looks amazing fam! Just tried it out for our company it generated amazing slides.
Very quick! Beautiful design too.

I cannot make particularly aesthetic pitch decks and slides but always love to look at good ones. Happy that I can create some too now!

Been loving it for my investor updates and all hands. Go Chronicle 🎖️

Just tried building out some presentations, fantastic tool! Love the real-time collaboration.

I like the idea of an AI design coworker. Instead of fully automating everything, it sounds like Chronicle helps you iterate and improve your story.

Congrats on the launch! :)

the mix of AI generation and manual editing makes this useful. I can quickly generate slides and still adjust everything they way I want before sharing it.

Congrats on the PH launch! The product quality improvement is so massive over last few months - excited to see what's next!

Watching this come together over the past months has been incredible. The team genuinely cares about making presentations better for everyone 👏

Excited to hunt Claude Code Review today! :)

As AI-generated code explodes, code review is becoming the bottleneck. Developers are shipping more code than ever, but PRs often get quick skims instead of deep reviews, letting subtle bugs slip into production.

Claude Code Review tackles this with a team of AI agents reviewing every pull request. Instead of one pass, multiple agents analyze the PR in parallel, verify potential issues, filter false positives, and rank bugs by severity.

What makes it interesting? It is the multi-agent architecture designed for depth over speed. The system scales reviews depending on PR complexity and leaves a high-signal summary plus inline bug comments directly in GitHub.

Key features

• Multi-agent PR reviews

• Parallel bug detection + verification

• Severity-ranked findings

• Inline GitHub comments

• Review depth scales with PR size

Benefits

• Catch bugs humans often miss

• Reduce reviewer workload

• Higher quality PR reviews

• More confidence when shipping AI-generated code

Who it’s for

Engineering teams, AI-heavy dev teams, and organizations managing large volumes of pull requests.

Use cases

• Reviewing AI-generated code

• Large refactors and complex PRs

• Security & logic bug detection

• Scaling code reviews across teams

Personally, I think this is a great example of agents solving real developer workflow bottlenecks, not just generating code but improving the quality of what gets shipped.

View details here:

• https://claude.com/blog/code-review

• https://code.claude.com/docs/en/code-review

What do you think? Share in the comments! :)

Who has a Team or Enterprise subscription?

The multi-agent review idea is interesting. AI can generate code fast, but reviewing it properly is still a challenge for many teams. Having multiple agents verify findings to reduce false positives sounds like a smart approach. Curious to see how it performs on large PRs.

Seems like Caude killed a lot of code review products from YC. They may have to pivot.

So we have AI writing the code, and now a team of AI agents reviewing the code. Are we humans just here to pay the AWS server bills now?Haha. Brilliant launch!

Huge launch, the multi-agent approach for PR reviews makes a lot of sense. Catching logic bugs, security issues, and subtle AI-generated code mistakes before production is exactly where teams need help.

Coincidentally, today I launched something related as well: Blocfeed.

While tools like Claude Code analyze the code itself, Blocfeed focuses on what happens after software reaches real users. Bugs often appear only on specific systems or edge cases where everything works fine on the developer’s machine.

Blocfeed aggregates user feedback and reports to surface:

• Bugs that only occur in certain environments

• Issues that slip past internal testing

• Patterns in what users are complaining about

• Feature requests users repeatedly ask for

I can imagine a strong synergy here:

Claude Code → prevents bugs before merge
Blocfeed → detects real-world issues and user needs after release

Congrats on the launch, excited to see where this multi-agent review direction goes. 🚀

Multi-agent review is exactly where code review needs to go. A single pass reviewer misses the same classes of bugs every time, but having specialized agents looking at security, logic, and performance in parallel catches the stuff that slips through. The false positive filtering is the make-or-break part though. Nothing kills developer trust in automated review faster than noisy findings they learn to ignore.

been building with Claude Code for months now and the "quick skim" problem is very real. agents write code fast but the subtle bugs pile up — especially when one agent changes something another agent built two weeks ago. multi-agent review makes a lot of sense here, curious how it handles context across larger PRs where the full picture only emerges from reading multiple files together.

This is honestly the missing piece for teams shipping fast with AI. I've seen so many PRs where the code "works" but has subtle auth bugs or logic holes that a human reviewer would catch on a good day but miss when reviewing 20 PRs.

The IDOR example in the demo is a perfect case. That exact bug pattern shows up constantly in AI-generated code because the model just focuses on making the endpoint functional, not secure. Having agents verify findings before flagging is smart too, cuts down on the noise.

Multi-agent code review is a great concept. Having different agents specialized for different types of issues — security, performance, logic errors — should catch things that a single-pass review would miss. Really like the approach of catching bugs early in AI-generated code specifically, since that is becoming the default way people write code now.

Congrats on the launch! Multi-agent review that verifies its own findings to reduce false positives is a nice touch. Noisy code review tools are worse than no tool at all. How are teams finding the signal-to-noise ratio so far in the research preview?

Built my entire SaaS with Claude Code, so this is relevant to me. The biggest challenge with AI-generated code isn't writing it, it's trusting it at scale. When you're integrating multiple ML models and wiring up payment flows, a missed edge case can cost you. Excited to see multi-agent review applied to this problem.

We started using Claude at the agency for client briefs and first-draft copy. The multi-agent review is a smart addition AI-generated code ships faster than anyone can review manually, so having agents check each other makes sense.Curious about the false positive rate. That's usually where automated review tools lose the team's trust.

want my team to switch from Greptile to Claude Code Review I Want few reasons especially for my CTO @raj_sharma_2000 cost comparison ?? Mermaid diagram

Hey Product Hunt 👋

I'm Jakub, founder of @Your Next Store. I've spent years in commerce, running my own software agency, then at Saleor, a platform powering enterprise brands like Lush and Breitling.

Those years taught me one thing: flexibility has a cost. Plugin hell, hidden fees, things that don't work together. Commerce has been a mess. So we started from scratch.

Instead of adding flexibility at every layer, we removed it. One payment provider, one clear model, no endless configuration. That's the whole idea behind Omakase Commerce. You don't assemble the stack. You trust the chef.

🎁 Product Hunt offer: Launch from PH and get your first month of the Starter Plan for $1.

Your Next Store is a Commerce OS built for agents and humans. The AI builder is your entry point: describe your store in plain English and get a real storefront connected to products, cart, and checkout from the first prompt. Not mockups. Not static pages.

Build like you would in Lovable - prompt, iterate, ship. Then agents take over: running the ops, surfacing actionable insights, and telling you exactly what to do next, so you can focus on growth and distribution.

Underneath is a full commerce layer built on our Commerce SDK. No plugin chaos, no five dashboards, and the freedom to change anything without breaking payments.

Your Next Store is how you become a 10x merchant - the AI builder gets you live in hours, the OS makes running your store feel effortless.

We'd love for you to check out some of our stores:
- rePebble by @ericmigi
- Sine Silk
- ChocoTales

Huge thanks to @chrismessina for hunting us 🙏

Congrats on the launch! Does you platform offer an in-built payments system?

There's something different in this AI builder - focused (e-commerce), opinionated (no plugins), and beautifully crafted.

S/O to @zaiste and team for the stunning work. The future of ecom is here.

Great launch! Congrats on being on the top of the leaderboard. By the way, what did you use to make the gallery images? They look so good.

Congrats on the launch! Is Your Next Store written on Next.JS? ;)

"Instead of adding flexibility, we removed it" is unironically the bravest and most beautiful pitch I've ever heard in e-commerce:D Death to plugin hell!

The Stripe-native + full code ownership angle is the part I'd actually lead with more — most agency teams I talk to are scared of AI tools that create lock-in or bury the payment logic in a black box. What I'm curious about is how it handles the store design side once the data model is sorted — because generating a working product catalog is one thing, but getting to something a client would actually put in front of customers is usually where the real time goes.

Checked how much you improved since the last launch. Amazing shipping velocity, congrats!

I’m curious how customizable the generated stores are after launch. Having full code ownership while still using AI workflows sounds like a really good balance.

Hey Jakub! It looks impressive and e-commerce game needed a refreshed face. Happy to see you helping on this. Quick question, can be used for drop-shipping?

This feels like a truly next-gen AI-powered no-code store builder for e-commerce owners. Congrats on the launch!

The AI-first approach to e-commerce is smart. Building a store used to mean choosing between no-code tools that are limiting or custom development that is expensive. Having AI handle the setup and customization while still giving you a real, production-ready store is a great middle ground. Congrats on the launch!

Great team, awesome product 👏 It's amazing to see you iterating at that pace. Brilliant tool.

Does it only work with Stripe for payments or can we add other payment methods too? Looks interesting, will share with my wife who is about opening a store now, perfect timing!

So if a store starts getting decent traffic, does it scale well? That's always been a pain with some platforms.

Congrats guys 👏

The “store in minutes” part caught my eye. How many minutes are we talking about realistically? 😄

Like from zero to first product live?

polish devs taking over ai space part 51231 👏

Really dig the approach here. An opinionated, Stripe-native commerce stack with full code ownership is exactly what agencies and dev teams actually want. The "built for the agentic future" angle with AI workflows plugging in through the API is smart. Any plans for a managed hosting option for non-technical founders?

Great work on this, guys! Congrats!

One of my biggest issues with Wix is the limitation to set up a way for customers to select delivery dates and times based upon our availability. For stores that require a bit more flexibility, will prompting via Your Next Store allow me to reach that degree of flexibility?

Love the MCP + agentic commerce direction. Curious — how does it handle multi-storefront setups for agencies managing multiple brands? That's usually where the plugin hell gets worst. Congrats on the launch!

explain plz to a non-tech small ecomm owner, how that'd be better than Shopify?

Feels like commerce stacks are slowly being rebuilt around the assumption that AI agents will operate the system, not just humans using dashboards.

The idea of exposing well-modeled primitives through APIs and letting agents orchestrate workflows on top makes a lot of sense.

Curious how you think about the boundary between AI-generated stores vs long-term maintainability and customization once companies start scaling?

@Your Next Store, Stripe-native approach is smart, removing payment abstraction layers reduces friction significantly. How do you handle multi-currency and tax compliance for international sellers compared to Shopify's built-in tools?

Really nice project! Congratulations on your launch!
Question: do you account for scalability in the design of the store as the volume potentially grows?

Great project and great execution! To the moon guys! 🚀 🌕

Great launch! Congrats on being on the top of the leaderboard!

Commerce rebuilt for the agentic future is the exact right framing. The cleanest way to think about it: the API-first stack is what makes a product agent-operable, not just AI-assisted. We're doing something similar in travel — the hard part isn't the LLM layer, it's making sure the primitives (availability, pricing, itinerary structure) are well-modeled enough that an agent can actually reason about them without hallucinating constraints. Curious how you handle the 'confirmation' problem: when a store built by an agent needs a human to approve something mid-flow, how do you surface that without breaking the whole experience?

Congrats on the launch! Curious how you’re thinking about integrations - will the platform stay centered around a smaller set of core tools, or open it up to a wider ecosystem over time?

Hi our beloved PH!
[excited] [slightly nervous]

Today we’re launching Fish Audio S2, our new text-to-speech model.

[long pause]

Hear Fish S2 Read This!

This is a big step beyond S1, redefining expressive voice AI. Write emotion cues anywhere in the text and hear the speech flow exactly how [emphasis] YOU direct it.

And, [inhale] we’re open-sourcing all of it.

GitHub: https://github.com/fishaudio/fish-speech/
HuggingFace: https://huggingface.co/fishaudio/s2-pro/

Shout out to SGLang for powering our stack.

There’s much more to S2.
Try it yourself now: https://fish.audio/s2/

As always, we want to give back to the community. For the launch, we’re offering free generation credits and an exclusive 50% OFF promo code: PH-FishS2

Go build weird things with it :)

We’d love to hear what you make.

How does Fish Audio maintain consistent emotional prosody and rhythmic nuance across long-form content, and what specific architectural improvements over So-VITS-SVC allow for such high-fidelity cloning from only 10 seconds of source audio?

big fish audio fans for a long time, been witness the team always go above and beyond. let's gooooo s2! congrats on this launch

Can I use this in a raspberri pi voice assistant that I have at home?
What abour the voice cloning to use it in phone calls?
eleven labs is not that good.. ( or I dont know how to set it up)

Excited to see the new version coming! Will it support any new languages?

Really cool how fast it can be to clone my voice. Should I be giving it multiple recordings at different emotions so that it has a better register of what I sound like?

exactly what we need, gonna try it now

Just found fish audio this year and was surprised about the API and the S1 model. Well, the S2 is now absolutely mind-blowing. Great work!

Oh my this is mind blowing. Does it support streaming on self hosted?

What's the basis for the tonation or emphasis? Congrats on the launch, @hehe6z!

Good job！

this is called gold mate! keep making more such products like these

This is a big unlock for anyone building voice-driven products. Directing voices with natural language cues like [whisper] or [laughing nervously] instead of fiddling with sliders is so much more intuitive. Love that it's open source too. What languages are you seeing the most community demand for?

As a content creator - I've been looking for a product like this for a long time! Hope it'll match my expectations.

Congrats on launching Fish Audio S2 expressive voice control with natural emotion tags looks very promising. Do the emotional tags also work well for Bengali and Hindi voices?

How do we leverage a multi-host generation in one pass? I can't find documentation on the site for producing the same level of quality that's portrayed in the demo. Could you point me in the right direction, please? Thank you so much. This demo looks really promising!

Also curious if you have a notebooklm alternative to the audio generation?

Hi Product Hunt! I'm Jochen, co-founder of sitefire (YC W26). 👋

My co-founder Vincent and I met at TU Munich, and have backgrounds in software engineering and reinforcement learning from Stanford. We started sitefire in late 2025 after becoming convinced of one thing:

Websites are going away. Going forward, people will interact with brands via AI agents like ChatGPT and OpenClaw. This means brands need to design their marketing content for AI agents, not just humans.

The problem? Most AI visibility tools stop at monitoring. They show you dashboards but don't help you actually do anything. We wanted to build the tool that takes action for you.

How sitefire works:

For every topic where you want to be visible, our AI agents analyze what content drives AI citations - top-cited pages, query fan-out, sourced domains, and more. Then, sitefire recommends one of four actions:

📝 Create content - Fully-written, brand-aware, AI-optimized articles based on top-cited pages, your sitemap, and SERP data. Push to your CMS (Framer, Webflow) in one click.

✨ Improve existing pages - Our AI agents know your sitemap. If you have content that could be cited but is not, we suggest tweaking it so LLMs are more likely to cite it.

📣 Earn media - See which PR outlets drive AI answers for your topics. Get research on how to approach them, incl. the publisher contact and email draft.

💬 Engage in communities - Find high-value Reddit threads and other forums that matter, with suggestions on what to post.

What's new today:

Starting today, every sitefire plan includes AI-optimized articles and one-click CMS publishing. We went from "here's what you should do" to "we did it for you.". This is our hello world moment.

Our product is free to try for 7 days. You can set up your account in 5 minutes and start getting your first content recommendations.

👉 We'd love your feedback: What's the biggest challenge you face with AI visibility for your brand?
Please tell us what is still wrong with our product. How can we make it better for you?

Thank you for your support! 🙏

This is awesome! I like not having to look at a dashboard and do the necessary changes myself when AI can do them for me :) Let's gooo

Your SEO agency just spent 6 months building 200 backlinks. Sitefire looked at the ChatGPT fanout query data and said "cute, but ChatGPT uses Bing and none of those pages rank there." The era of vibes-based SEO is cooked. 🫡

Great stuff! :) Supported and shared in our internal channels. Best of luck!

How does sitefire identify the specific content features—such as structural patterns, entity relationships, or citation density—that successfully trigger citations from diverse AI answer engines like ChatGPT, Gemini, and Perplexity?

Can you tell us more about the roadmap you're planning for the next few months?

@jochenmadler congrats on the launch!

Congratulations!!!

Strong product, we have been looking for the right AEO solution for a while and Jochen & Vincent have built something amazing! They are also super responsive to feature requests and we have already had significant improvements in our traffic, couple weeks since using sitefire

That looks awesome. Congrats on the launch!

Interesting launch, @jochenmadler! Congrats to you and @vincent_jeltsch1.

What stood out to me is the loop you’re closing. Sitefire analyzes what gets cited. Then generates brand-aware content. Then pushes it directly to the CMS.

That “research → content → publish” flow is powerful.

One question while reading through the page.

When your agents analyze citations across models like ChatGPT, Gemini, and Perplexity, do you see different citation patterns between them?

Or do the same types of sources tend to appear across models?

I'm excited to see how this evolves. Great launch.

The whole "take action, not just monitor" angle is really solid. Most tools in this space just give you dashboards and leave you to figure out the next step yourself. Pushing directly to Framer/Webflow is a nice touch.

Congrats on the launch! One thing that might help as you scale, have you thought about adding an in-app feedback widget? Something like Blocfeed where your users can report issues or suggest features right from inside the app. Helps you understand what's actually breaking for people in prod and what they want next. Could be useful for prioritizing your roadmap early on.

The distribution layer of the internet is changing so fast it’s hard to keep up. There are already dozens of tools giving analytics, but I rarely see them turning those insights into action.

Super curious how often and how you have to reverse engineer what the models are actually doing.
Congrats on the launch!

The thesis that brands need to optimize for AI agents, not just humans, is really compelling. Most marketing tools are still stuck in the SEO-for-Google mindset while the landscape is shifting fast toward LLM-driven discovery. The fact that sitefire actually takes action (creating content, improving pages) instead of just showing dashboards is a big deal. Smart move going through YC with this timing.

Sitefire is a great way to get your brand discovered! So excited for this launch!

@jochenmadler @vincent_jeltsch1 congrats on the launch 🚀 curious - you are writing that you are helping with earned media, how specifically does you product help with that?

The citation piece is really interesting. With the content creation piece, is this meant to serve as a "reviewer" that helps make recommendations on where to make improvements in order to get a higher chance of being cited by AI? Or is this something that writes content for me?

This is epiq! How do you track the impact of Sitefire?

The thesis that brands need to optimize for AI agents and not just humans is going to age well. Most marketing teams are still thinking about this as an SEO problem when it's really a completely different distribution channel. The action-oriented approach over dashboards is the right call. Nobody needs another monitoring tool that tells them they have a problem without fixing it. How are you thinking about the feedback loop when LLMs update their citation behavior? What works today might not work in 3 months.

Curious how you handle the quality control loop when agents push content directly to a CMS — that's where I'd get nervous. Auto-publishing brand-aware articles sounds good until one goes out that's slightly off-tone and you're doing damage control. The citation analysis piece is the most interesting part to me, because understanding what content actually influences AI answers is still pretty murky for most teams. We've been thinking about this at told.club from a different angle — what users say in feedback often ends up being the raw material that shapes how a brand gets described, and that gap between company-published content and user language is huge. Would love to know if you're pulling in any of that signal or just working from existing indexed content.

Exactly what I was looking for, congrats on the launch @jochenmadler and @vincent_jeltsch1. How do I get more of our content into AI search?

Great work! Congrats on the launch!

Quick question: Would you say Sitefire already makes sense for smaller or low-authority websites, or is it mainly useful once a site already has some domain authority?

Hey Product Hunt community!

Today is our 4th Product Hunt launch overall and the first one after the acquisition.

Since the beginning of the year we’ve been moving fast. Besides improving product performance, we’re focusing heavily on expanding the agent’s skills and what it can do. Because we believe when the agent can do more, you can build better things.

While watching people use Macaly we noticed something interesting. Users sometimes try a few obvious prompts but don't go deep. Not because the agent is limited, but because they don’t know what to ask it to build.

So we decided to show it.

Today we’re sharing 15 things you can ask the Macaly agent to build:

• Turn a YouTube video into a landing page

• Redesign a website from a URL

• Turn a spreadsheet or PDF into pages and dashboards

• Generate a database for your app

• Add login, sign-up and user roles

• Build dashboards and admin panels

• Add AI features like chatbots or smart search

• Match the style of a screenshot or design reference

• Create forms that store submissions in a database

• Send automated emails like confirmations or alerts

…and a few more.

Have an idea for a new skill? Share it in the comments. If we like it, we’ll reward you.

We also added a special launch coupon for 33% off your first three months. And there’s a way to get extra credits on the new landing page if you look closely.

Freshly joined the Macaly crew — ready for the day 1 feedback.

Bugs, feature requests, "this is genius/miss" — hit me!

Wow! Really cool guys)

Does the "add AI features" capability include semantic

search or is it limited to basic AI calls? Also,

can generated apps connect to existing databases or only use Macaly's

built-in ones?

Congrats on the launch!

The landing page from YouTube video feature caught my eye — that's a creative use case.

How accurate is the output?

Hey folks, you rock. I love the ease when creating stuff on Macaly. In the beginning i remember there were some problems with stability, but those are solved. I also loved the speed of your development in so few people. I keep my fingers crossed for you.

i'm using Macaly for a month now and I'm really happy with what it can do. My only issue is when it creates bugs and then it fixes these bugs by using credits. that's not cool.

Congratulations on the fourth launch, Macaly team!

Good luck guys!

Turn a YouTube video into a landing page? How is this even real? That's amazing! Congrats on the launch, @petrbrzek!

The discovery problem is real — most people don't know what to ask, so they never push the tool past the basics. Showing 15 concrete use cases is probably more useful than any onboarding flow. The part I'm curious about is how you handle the gap between what the agent can do in demo mode versus what actually works in someone's specific stack — that's usually where expectations break. Does it adapt to context or is it more of a fixed menu?

Are your website designs better than others?

Hey Product Hunt 👋,

Ashwin here, co-founder of Spine.

Spine lets you manage a team of AI agents that work together to complete complex tasks: researching, analyzing, and building full deliverables like apps, landing pages, documents, spreadsheets, presentations. All on one visual canvas you can watch in real time.

Instead of one model doing everything, Spine spins up specialized agents in parallel, picking from 300+ models to use the best one for each step. The result is finished deliverables — not a chat response.

A good first task: "Research [your industry] and create a competitive analysis with a market map, executive summary, and strategic recommendations."

You'll see multiple AI agents spin up simultaneously, browsing the web, structuring data, assembling deliverables. For large projects it can run autonomously for 80+ minutes.

It's free to get started, just signup, no terminal or installation needed. Excited to see what you build 🚀.

Drop your results in the comments — we're reading everything today.

Hey 👋, Akshay here, co-founder of Spine.

If you've seen what OpenClaw and Claude Code can do for developers — autonomous agents running for hours, finishing real work — Spine brings that same power to everyone. No terminal, no setup. You describe what you need, and a team of agents executes it on a visual canvas.

We just scored 87.6% on Google DeepMind's DeepSearchQA (this measures how well AI answers complex research questions) — ahead of Perplexity (79.5%), Claude Opus 4.5 (76.1%), GPT-5.2 (71.3%), and OpenAI Deep Research (44.2%). We're 8 people. Turns out the right team of agents, drawn from 300+ models, working together on a canvas beats any single model working alone.

Some tasks worth trying:

→ Ask it to audit your website and produce a growth roadmap with a full slide deck.
→ Preparing for a fundraise? Give it your company details and get back a pitch deck, competitive landscape with market sizing, a financial model, and personalized outreach emails for target investors.

→ Describe a product idea and get back multiple interactive prototypes and landing pages exploring different design directions, alongside a PRD and go-to-market strategy.

Everything Spine builds — docs, spreadsheets, decks, prototypes, landing pages — is downloadable and hosted at a shareable link. Just send your team the URL.

Product Hunt exclusive: use code PHLAUNCH10 for 10% off any paid plan.

Would love to hear what the community builds with it!

Let's go!! 🚀 I used Spine with a client the other day who gave me a disjointed mess of documents and links and said "design me a website that's like these!"

I fed everything into Spine before bed and woke up with mockups, a comprehensive design guide, full decision making process, and a report I could hand the client as to the direction I was going & the next steps. And... it did better than I ever could.

1 day of work saved thanks to Spine 💪

Working on Spine made me realise how different things get once you move from one model to many agents. A lot of the engineering ends up being orchestration routing tasks across models, coordinating long-running jobs, and keeping outputs structured so other agents can build on them.

It’s interesting watching it break down larger tasks and assemble real outputs research reports, strategy docs, prototypes, landing pages, and slide decks.

I’ve also been using it for engineering workflows like researching systems, summarising docs and blogs for quick reads, getting perspectives from multiple agents with different personas, generating quick prototypes, and thinking through edge cases.

Would love to see people try workflows like this too.

Pretty fun system to build and work on.

Real productivity platform where AI agents help me complete my work. Love this product!!

Engineer on the team here,

One of my favorite moments while building Spine was the first time we ran a task and just… watched agents work for ~30 minutes assembling a full research doc and slide deck.

With Spine Swarm, it felt less like prompting an AI and more like assigning a project to a small team who do their job incredibly good.

I'm on the team at Spine but I also use it constantly for my own work. What I keep coming back to is how visual everything is — agents do the heavy lifting, and you get real deliverables on a canvas you can actually look through, rearrange, and build on. If you've ever wished you could just watch AI work and step in when it matters, that's basically this.

Love it! Can't wait to try this out on marketing ops.
Supported and shared on our channels. :) Best of luck!

The idea of spinning up specialized agents in parallel and picking the best model for each step is really smart. Most AI tools try to do everything with one model, which leads to mediocre results across the board. The visual canvas where you can watch agents work in real time is a nice touch too — transparency in how AI arrives at deliverables builds a lot of trust.

Curious how the swarm coordination actually works when agents hit conflicting conclusions mid-task — like if one agent's research contradicts another's during a 50-page strategy doc. That's where these multi-agent setups tend to fall apart in my experience. The visual canvas angle is smart though, auditability is genuinely the missing piece in most AI workflows right now. Most people don't trust the output because they can't see how it got there.

Canvas over chat - that just makes more sense for real work. When I'm jumping between research, code, and product decisions, linear chat loses context fast. Love the branching idea. Quick question - can you connect outputs between blocks automatically or is it all manual?

Interesting architecture. Orchestrating multiple specialized agents across 300+ models to decompose long running tasks into structured outputs on a shared canvas is a strong systems design choice.

Curious how you handle task routing, intermediate state management, and verification of outputs between agents to maintain consistency.

I've watched people use this for the first time and the moment it clicks is always the same, they don't just prompt and wait as AI goes behind a black box to stitch up an answer, but prompt and watch as their specialised AI workforce comes together in parallel to deliver real work, live.

Couldn't be prouder of what this team built. Go try it!

I have tried Spine Swarm and it seems as v.useful tool in the AI toolbox.

This is pretty good!

Interesting system design here.

From the description it feels like Spine behaves closer to an orchestration layer coordinating swarms of agents rather than just a typical AI workspace.

Curious how the team internally thinks about that distinction.

Hey everyone! Maker here.

I've been using Claude Code skills for months and they're genuinely incredible.
For me, they're like the scene in The Matrix where Neo gets Kung Fu uploaded directly into his brain.

A single SKILL.md file can completely transform how your agent works and make it learn about SEO, how to write cold emails or even how to do accounting in France.

However finding good ones? Total mess. Scattered across thousands of GitHub repos with no way to search or compare. Then the ClawHub malware incident happened. 20% of submitted skills were malicious. Prompt injection, credential theft, obfuscated code.

So I built agentskill.sh that currently indexes 100k+ skills for Claude Code, Cursor, Codex, Windsurf and more; I focused on two things:

1. Security: Every skill gets scanned across 12 threat categories so you know what you're installing before you install it. You can check the details here: Security Dashboard

2. Discovery: You can search skills by categories using many criteria, review them, and more.
   

The fastest way to try it is the /learn command. Once installed you (or your agent) can search and install skills directly using:

/learn               # just find skills for current codebase

/learn seo           # search by keyword

/learn @owner/name   # install a specific skill

/learn trending      # see what's popular

Using /learn to find skills has another big advantage: it lets your agent learn by itself.

When your agent hits a problem it doesn't know how to solve, it can search for and install the right skill on its own. No manual hunting, no copy pasting from GitHub. Your agent just gets smarter as it works.

What makes /learn special:

1. Two layer security. Every skill on agentskill.sh is scanned server side for 12 threat categories (command injection, data exfiltration, credential harvesting...). Then /learn performs a second client side verification before installing. You get both centralized scanning and local confirmation.

2. Feedback loop. Your agent auto rates skills after using them, so the best ones surface and broken ones get flagged by the community. Your agent contributes to, and benefits from, collective quality signals.

3. No context switch. Search 100k+ skills mid conversation, install what you need, and keep working.

Would love to hear what skills you're using and have your feedback on this!

The discovery problem for skills is real — right now it's mostly vibes and GitHub spelunking. Having a searchable index with some vetting behind it is exactly what this space needs. Congrats on the launch!

Cool graphics!

I spend a serious amount of time hunting down and wiring up Claude Code skills for our accounting automation stack — discoverability is the real friction. Most skill repos are scattered or undocumented. Centralizing this is the right call. Curious: do you curate for quality before listing, and is there a way to pin to a specific skill version so a production workflow doesn't break when a skill gets updated?

Installed and testing.  If I understand correctly, the brilliant part of the /learn implementation is that it has context-awareness built in. If you just type /learn into the chat with no other context, the skill instructs the agent to look at the environment (e.g., scan package.json, check file extensions, look at the current Git branch name), then automatically query the database and say:

"Based on your project, I see you are using Next.js and Prisma. I recommend installing the nextjs-app-router and prisma-schema-expert skills. Would you like me to install them?"

Is that correct?

The security scanning is the real differentiator here. 20% malicious rate on submitted skills is wild but not surprising given how the ecosystem exploded. Discovery is the other half that matters. Right now finding good skills is like searching GitHub in 2010. Having a curated, searchable index with security scores changes the game for teams that want to adopt agent skills without rolling the dice on supply chain attacks.

If this helps make skill building and selection even 10% easier it's worth its weight in gold!
Happy to support - best of luck!

Handy!

Thank you @romainsimon , lots of skills I didn't know!

The URL ends up in a 404, you may want fix the URL as it contains "@"

https://agentskill.sh/@?ref=producthunt

Hey Product Hunters 🫶

I'm Dimittri, co-founder of Sonarly (YC W26)! Excited to introduce Sonarly, the AI that fixes your software autonomously.

More code than ever is being pushed, which means more bugs than ever arrive in production.
When production breaks, you don't see one alert, you see an avalanche. Sentry, Datadog and friends all light up at once. Some alerts are critical, some are duplicates, some are just noise. You click through them one by one, jump to logs, try to piece together a root cause, and end up wondering if your alerting is even worth paying attention to.

Sonarly solves all of that.

We bridge the gap between monitoring tools and coding agents to finally make software improve itself.

Powered by Claude Code and Opus 4.6, Sonarly is the most powerful AI agent to deduplicate alerts and fix bugs with an optimized context of your production system, including access to your codebase, logs, metrics, and traces.

Connect your monitoring tools and watch Sonarly investigate and deduplicate while you focus on shipping and talking to your customers!

What you get:

- An intelligent grouping to remove the noise and duplicates from your alerts

- Automatic investigation and fixes with PRs and evidence to ensure nothing is hallucinated

- Slack-native integration to fit directly into your workflow

We started Sonarly because we've been building products since we were 16, but were frustrated that no tool existed to bridge monitoring and fixing, it felt so obvious to us.

We work with fast-growing companies and help them cut their resolution time, and now we want to help you do the same.

We built the tool we wish we had when we started. Sonarly is the tool we dreamed about when we got our first users.

Try Sonarly at sonarly.com with 100% off for next 2 weeks, promo code SONARLYHN (only for Hunters!)

Huge fan of this product it will change the way people do monitoring!

Curious how you handle the cases where the AI-generated PR fixes the symptom but not the root cause — that seems like the failure mode that would erode trust the fastest. The monitoring + auto-fix loop is interesting but it requires a lot of trust in the system, and most eng teams I know are still pretty cautious about auto-merging anything. What does the human review step actually look like in practice?

The "evidence to ensure nothing is hallucinated" framing is doing a lot of work in your pitch. What does that actually look like in the PR? Are we talking linked log lines, specific stack traces, a confidence score on the root cause?

amazing product and team!

Really cool that you're using Claude Code under the hood for this. The alert deduplication alone sounds like it would save a ton of time. Most teams I know just drown in Sentry noise and end up ignoring half of it.

One thing that pairs well with backend monitoring like this is collecting bug reports from actual users too. Something like Blocfeed lets users click on the broken element and submit reports with full context (CSS selectors, console errors, browser info). Helps you catch the stuff that monitoring tools miss because users experience things differently.

Congrats on the YC launch! How does it handle false positives? Like if Sentry fires an alert that's not actually a real issue, does Sonarly still try to "fix" it?

Been using @Sonarly for a while and big fan. Really helps to cut through all the emails I get from Sentry. Spending much less time on alerts and fixes now. Thanks for building this @dimittri @alexandre_k_

Demo video? Need to see it live.

Well done team!

Sounds useful. Happy to support! :)

The hardest design question for autonomous agents in production isn't technical accuracy — it's calibrating when to act vs. when to ask.

Too conservative: you're just a smarter alert tool. Too aggressive: one bad autonomous fix and engineers stop trusting the whole system. The trust ratchet cuts both ways.

Building an AI travel planner I ran into the same boundary: full itinerary autonomy felt threatening until the AI had earned credibility on smaller decisions first. The graduated handoff — suggest, then draft, then execute — isn't just UX, it's how you build the confidence that makes full autonomy acceptable.

Curious how you're thinking about the autonomy spectrum on Sonarly. Is the "fix autonomously" behavior configurable per team, repo, or severity level? Or is the current bet that the root cause accuracy is high enough to go straight to full auto?

Interesting architecture here.

From the description it feels like Sonarly behaves closer to an autonomous incident-response layer rather than just a typical monitoring or debugging tool.

Curious how the team internally thinks about that distinction.

Met the team, they are sooo good! Shared with all the CTOs in my Network

How's the false positive rate on PRs?

How does Cowork manage state and error recovery for multi-step tasks that run for hours, ensuring that "Work IQ" remains synchronized if relevant documents or context change during the execution process?

Most AI tools are great at the first 20%. They find the answer, draft the paragraph, surface the insight. Then they hand it back to you and the real work begins. That handoff is where hours disappear.

Copilot Cowork is Microsoft's attempt to close that gap. It's not a chat upgrade, it's a task-execution layer built into M365. You describe the outcome you want, and it turns that into a plan that runs across your actual apps and files.

The underlying insight here is important: intent and execution have always been separated in software. You tell the tool what you want. The tool tells you what it found. You do the rest. Cowork collapses that loop.

What it actually does:

• Calendar triage 📅 -- reviews your Outlook schedule, flags conflicts, proposes changes, and applies them once you approve

• Meeting prep 📋 -- pulls inputs from email, files, and past meetings, then produces a briefing doc, a deck, and a draft follow-up in one pass

• Company research 🔍 -- pulls earnings, SEC filings, analyst coverage, and news, then outputs an exec summary, structured memo, and a labeled Excel workbook

• Launch planning 🚀 -- builds competitive intel in Excel, drafts a value prop doc, and generates a pitch deck -- coordinated, not siloed

Every action is auditable, runs within M365's security boundaries, and requires your approval before it ships. You stay in control. It just does the legwork.

Worth noting: Microsoft built this in partnership with Anthropic, with Claude powering parts of the execution layer.

That multi-model architecture is genuinely interesting. It suggests the future of enterprise AI isn't one model winning, it's the right model for the right job.

Who this is for: M365 users, knowledge workers, ops teams, and founders using Microsoft's stack who are tired of AI that helps them think but doesn't help them ship.

It's currently in Research Preview. Broader access rolls out through Microsoft's Frontier program in late March.

As AI moves from copilot to co-worker, where do you actually want to hand off control, and where do you want to stay in the driver's seat?

Funny launch of you @rohanrecommends looking forward to Thursday!

Hi Product Hunt 👋

I’m Stan, the maker of humansfix.ai.

Over the past year AI tools like Lovable, Replit, Cursor and v0 made it possible for non-technical people to build real apps. That’s amazing to see.

But there’s a moment many builders hit - something breaks and they don’t know how to fix it. Payments fail, deployments crash, integrations stop working, or the AI-generated code becomes hard to understand.

I built humansfix.ai to help in that moment.

The idea is simple: connect non-technical builders with real developers who can help with AI-built apps. Instead of hiring someone or paying hourly, you can post a task, set the price, and a developer picks it up. Results are delivered within a few days and you only pay after approving the result.

I’m curious what you think.

Thanks for checking it out 🙏

Curious how you handle the matching — does the developer self-select based on price, or is there some kind of routing on your end? The gap between "AI built it" and "I can actually maintain it" is very real, and I've seen a lot of teams hit a wall three weeks after shipping something on Lovable. Fixed pricing is smart here because the ambiguity of hourly rates would kill trust with non-technical users. The real challenge is probably scope creep — someone describes a "small bug" and it turns out the whole data model is broken.

Love the concept, I think it answers a very relevant nieche. supported and spread in our internal channels. :)

Stellar idea guys) It is like the next 'StackOverFlow-Generation') but cooler)

Do developers also help with code review and optimization,

or just bug fixes? And what's the average price range for typical tasks

like fixing a broken API integration or adding authentication?

This is very much needed. I know quite a few people people who built something cool with Lovable and then got completely stuck when payments broke or deployment just died on them. The "set your own price" model is interesting too. How do you make sure the devs are actually good though? Any review system or do you vet them before they join?

This hits a real pain point. I've seen so many people build cool stuff with Cursor and Lovable but then have no idea what to do when something breaks. Having a marketplace to connect them with actual devs is smart.

One thing that could pair really well with this is giving those non-technical builders a way to actually capture what's broken. Something like Blocfeed lets users click on the exact element that's not working and it grabs all the technical context (console errors, CSS selectors, browser info). Would make it way easier for the developers on your platform to understand and fix the issue faster.

How are you handling the matching between builders and devs? Is it based on tech stack, or more like a general queue where anyone can pick up a task?

How does HumansFix.ai manage the handoff of AI-generated codebases between non-technical users and developers to ensure that custom fixes don't break the original "vibecoding" workflow or compatibility with tools like Lovable and Cursor?

I think that this is a cool complementary website to PH. Also like @ProblemHunt :)

really like the idea, although more sorting options would be very helpful. Besides, some filters that target the most painful points not from big players but mid-size apps

Smart angle — mining 1-star reviews is something I've done manually for competitor analysis, but it's brutally tedious at scale. Does GapHunt currently only pull from iOS App Store, or are Google Play and web SaaS reviews (G2, Trustpilot) on the roadmap? That's where a lot of B2B pain lives.

Good and useful for problem identification and the first steps to start building a startup. I'd advise promoting the tool in pre-accelerators, startup schools and so on

As a non-dev founder, I’m perpetually worried about building the 'wrong' thing. Gaphunt sounds like exactly the reality check I need. Does it currently support all App Store categories, or is it focused on specific niches? Love the 'goldmine' analogy—super clever!

Really smart angle here. Mining bad reviews for product ideas is honestly something more founders should be doing instead of guessing what to build next.

• Love that you're pulling from 1 and 2 star reviews specifically. That's where the real pain points live.

• The AI analysis layer on top is a nice touch. Manually reading through hundreds of reviews gets old fast.

• One thought: once someone builds the app they find through GapHunt, they'll need a way to collect feedback from their own users too. Something like Blocfeed (in-app bug reporting widget) could help close that loop and keep validating with real user input.

Congrats on the launch, this is a solid tool for the ideation phase!

Mining bad reviews for product gaps is brilliant. I spent months manually going through competitor reviews before building my last project. The patterns you find in 1-star reviews often reveal the biggest market opportunities that everyone else is missing.

Hey Product Hunt, I’m Mike, founder of Refero.

We built Refero MCP because most AI-generated interfaces still look generic. Models are great at code and logic, but product design is a different skill. They usually don’t know which patterns work, how real products structure flows, or what makes an interface feel thoughtful.

Refero MCP gives AI agents access to 125,000+ real product screens and 8,000+ user flows, so instead of designing in a vacuum, they can study real products before generating UI.

The idea is simple: better references lead to better design output.
Less guessing, less generic UI, and more interfaces that actually feel designed.

Try it live: https://demo.refero.design/ or integrate into your agent https://doc.refero.design/mcp/getting-started

Would love your feedback <3

Tried this with a "node-based editor" prompt and the results were solid

couple of small things:
- explore next suggestions might be buttons so i don't have to copy-paste it
- downloadable report would be very useful. like a PDF or .md report that i can pass to the agents.
- to be able to pass my existing UI or website and get suggestions would be cool

Great tool, planning to use it later again!

So cool! Was looking for this when writing a report, does it work for agents that help in ppt/doc design?

Played around with the demo and it is super useful. If only for UI research. Very cool! shared it with our PM. Best of luck!

Yo Refero team!!

I’m a researcher for the H1Gallery newsletters (you can google us). We highlight excellent homepage headlines from interesting startups around the web.

We’re featuring Refero in our March 13th issue- the headline “Give your AI agent design taste” really stood out to us. Great positioning and clarity with the copy.

Would love to include a quick quote from someone on the team about the copywriting strategy behind the headline and how you approached the messaging. Totally optional of course, we’re featuring it regardless, but it’s always great for readers to hear the thinking behind the copy.

If you’re up for it, just send over a couple sentences and we’ll include it. Appreicate your time thank you :)

Best of luck! Shared with out PMs internally. :)

This solves something I deal with daily. The GitHub repo mapping is really interesting for teams like ours where the codebase is growing fast and AI tools keep losing track of the bigger picture. Just a question - when the codebase evolves over time, does the mapping update automatically or do you need to re-run it manually to keep things in sync?

How does CodeGuide maintain a real-time, bidirectional mapping of an existing GitHub codebase to ensure that the generated PRDs and tech stacks remain synchronized as the underlying repository evolves during an active development cycle?

This is super useful. I've been giving Cursor raw ideas and watching it go in circles because there's no structured spec to work from. Having PRDs and wireframes that AI tools can actually parse feels like a missing piece in the whole vibe coding workflow.

Congrats on the launch! Since you're building a web app, you might wanna check out Blocfeed too. It's a free in-app feedback widget that lets your users click on any element and report bugs with full technical context. Would help you understand what's breaking for people and what they actually want next.

How does it handle existing codebases that are already messy? Like if you map a repo that has inconsistent naming and no clear architecture, does it still generate useful specs?

As a developer who spends half my time in Cursor, I've noticed that 'vibe coding' only gets you so far before things start to drift. This structured context layer seems like a massive DX improvement! Being able to map existing GitHub codebases to feed into the AI is exactly what's been missing. Really excited to see how this reduces hallucinations in complex user flows. Great job on the launch!

Smart idea! Does it intelligently avoid quitting system apps or apps with active processes (like downloads)?

And can you whitelist certain apps to never auto-quit?

How does MacQuit handle background helper processes or menu-bar-only apps that don’t have a standard window state when executing a "Quit All" command, and what specific criteria does the idle timer use to differentiate between an inactive app and one performing a background task like rendering or syncing?

Hi Product Hunt! 👋

I built Book Reading Habit to help people read more consistently, starting with something simple: short reading sessions that feel easy enough to stick with.

The idea came from my own experience. I started with short reading sessions — just 5 to 10 minutes at a time — because it felt easy enough to stick with. Over time, I gradually increased the session length, and that small habit worked surprisingly well. Eventually, I wanted an app built around that experience, so I made one.

Book Reading Habit is designed to make reading feel simple, motivating, and sustainable. It currently includes features like:

• Build a reading habit with short sessions, flexible timers, and reading progress tracking

• Capture notes whenever you want — while reading or later

• Organize books with custom shelves

• Use widgets and Siri Shortcuts for quicker access

• Import and export your library

One of the biggest priorities while building it was privacy.

A lot of apps today depend on ads or social features that turn everything into a public activity. I wanted the opposite: a calmer, more personal reading experience. So Book Reading Habit is built to help you focus on reading without ads, unnecessary noise, or social pressure.

The app has a free version that lets you add up to 5 books, with unlimited sessions and shelves. If you want more, there are monthly, yearly, and lifetime upgrade options that unlock features like import/export, automations, unlimited books, and more.

I’d love your feedback:

• What feels most useful?

• What feels missing?

As goodreads remains one of the biggest players on the market, can the data from there be imported/synced or would hve to manually insert it?

I love this idea, I know so many individuals out there that have such a large TBR stack that never seems to dwindle. I also can see this help non readers get into reading by starting slow and learning to enjoy it versus getting forced into reading.

This launch reminds me there are so many books I've purchased which I've not read. I should start reading.

Short sessions are exactly what works — starting with 10 minutes and gradually increasing is much more sustainable than telling yourself you will read an hour a day. Curious how the notes work. Can they be linked to a specific page or chapter or is it more of a general notebook per book?

This will save a lot of time for me, I don't have to wait for every 20s ads before i can read what i need.

How does Shipper handle the transition from a "live business" back to a standard development environment if a user needs to export their project, and what specific infrastructure does it automate—such as payment gateways, authentication, or hosting—to ensure the app is truly functional from the first prompt?

Hey PH! David here, co-founder at Shipper 😸 👋

We made an AI agent that brings out the entrepreneur in anyone. It builds web/mobile apps, websites, Chrome extensions & more.

The biggest update from Shipper 1.0 is the Build Queue (picture #2): you can now see exactly what’s happening when you give Shipper a task, step by step: what’s being generated, what’s being configured, what’s being deployed.

_ _ _ _ _ _ _ _

Most of what's new in Shipper 2.0:

• Shipper builds anything now: websites, mobile apps, web apps, Chrome extensions, eCommerce stores etc

• New layout with a sidebar menu
• Build Queue with stacked prompts + step-by-step transparency
• The Shipper Advisor's roadmap planning
• App Store asset generation (images, descriptions, privacy policies)
• Animated websites powered by Google Veo
• Built-in video generation via Remotion
• One-click native app translation in 68 languages
• Native revenue-focused analytics dashboard
• Co-op building with shared workspaces
• Deep connectors: Stripe, Google Workspace, ChatGPT, Shopify, ElevenLabs & more
• Full version history and 1-click rollback
• Faster custom domain setup

• In-house domain purchasing with special deals, with domains as low as $1
• 50% commission + lifetime benefits affiliate program

• ... and +27 other smaller updates!

Thanks a lot for showing interest in Shipper!

Happy to answer any questions in the comments :)

Best,

David, Daniel & Team Shipper

Huge fan of anything that helps ship faster! As someone who often gets stuck in the 'dev loop' and forgets about the actual launch part, this kind of streamlined workflow is exactly what I need for my side projects. Quick question: Does it handle multiple platform deployments (e.g. PH, Twitter, etc) from one dashboard or is it more focused on the prep?