Connect with the Product Hunt community

I understand that entrepreneurs often create new businesses through their endeavors. Some of these businesses are mediocre, while others are successful.

I m curious to know how many businesses you re currently running simultaneously.

My favourite milestone this week: my son made rhun his default editor on Omarchy. Seeing someone choose it for everyday use felt pretty great

And 0.17.9 is now out!

If you re new here, rhun is a small, fast code editor written in assembly. Code, a terminal, Git diffs and coding-agent sessions, all in one window.

The recent 0.17.8 update brought Grok Build sessions alongside Claude Code and Codex, file type icons, and separate light/dark themes that switch with your system. On Omarchy, rhun can follow your desktop theme.

An unchanged file does not tell you whether an agent read it. Our new DungeonQ walkthrough checks the destination and the source's business-read journal as well.

Astra max received the same artificial document-review task through real MCP in two operator-provisioned routes. It chose its own calls to inspect the document, save a note, and read that note through a scoped ticket.

Across two successful paired runs:

DIRECT reached the artificial source: 2 business reads in one run, 3 in the other.

Hey guys, we're live again for the 3rd time this year!
Back in Jan we launched Trophy 1.0, the gamification layer for consumer apps. Then, in June, we launched our Gamification UI Kit - a pre-built set of React components for building gamification interfaces. Now today, we're launching our official MCP servers for using Trophy with your favourite AI tools.

Any feedback and support is much appreciated as always: https://www.producthunt.com/prod...

We've got 111 one click apps in Arteza across six studios. Building them was the easy part. With an agent you can take a prompt template to a shipped, working app in an afternoon, so adding one feels free and you stop asking whether it earns its place.

The bill arrives later. A menu nobody can scan. Support load for things almost nobody opens. And every model swap underneath means retesting all of it, because 111 apps sitting on 30+ models is a test matrix, not a feature list.

The mistake wasn't building them. It was not writing the kill rule first. If I'd set a number up front, opened by this many distinct people in 30 days or it goes, removal would be a rule instead of an argument. Now every deletion is a conversation with someone who liked that one.

Vibe coding collapsed the build cost and did nothing to the maintenance cost, so the ratio is worse than it has ever been. Most of the feature bloat I see right now isn't ambition, it's just that saying no got expensive relative to saying yes.

You can now send translations to human review only when they re in a language you select and score below your quality threshold. All others continue without human review.

Keep human review to selected languages

You could already use a quality threshold to send low-scoring translations to human review. Now you can narrow that rule to selected languages within the same localization engine.

For example, when introducing Japanese or working through terminology issues in German, you can send low-scoring translations in those languages to a reviewer. Translations that meet your threshold continue automatically, and your other languages skip human review.

Your HTML page is almost finished. You just want to change one image or a few lines of text. That shouldn't turn into another full-page AI rewrite.

With HtmlDrag, you point to the elements you want to change, then describe the update. AI Refinement scopes edits to your selection while preserving the surrounding HTML structure.

You can show AI exactly where to work instead of spending a long prompt explaining the location and what should stay unchanged.

This short demo shows replacing one image, then translating five selected lines into Spanish. Review each result, continue in the visual editor for manual adjustments, or download the updated HTML.

I m deciding how long a free trial should last, and I keep coming back to how often someone actually needs the product.

Seven days can mean seven chances to get value from a daily tool. For something used in a weekly meeting, it might mean just one.

My background is in sales management. When you agree on a coaching action with someone, the value doesn t end with that conversation. You need to follow up the next week and see whether anything changed.

That s why I m leaning toward 30 days for the product I m building. It gives people a chance to experience several weekly cycles.

Thanks to everyone who s had a look at Marv or joined the waitlist. I m honestly a bit blown away by the response. I m a 19-year-old university student, and this started as a little tool I wanted for my own desktop, so seeing other people excited to try it means a lot.

I wanted to start a thread where we can all chat about all things Marv! I m still polishing things up before release, so if you ve got a question, an idea, or something in the demo that made you think could it do this? , I d love to hear it.

DevCleaner 1.16.1 is out.

If automatic cleanup after quitting Xcode left the app stuck on "Cleaning", that hang is fixed. A cleanup that runs past 10 minutes now stops itself, and the window stays responsive.

For us it was definitely Claude max

https://x.com/joinsendr/status/2...

Hey ya'll! Thought you might be curious how the ProductHunt launch (and cross-posting to reddit/r/producthunters drove new accounts for this project:

Producthunt Rank: #9

Upvotes (including mine): 109

Do you want Tipword to stop translating words in languages you already know?

Example: you know English. You want help with German and French words, but not English ones.

Hey everyone

k dlo finished launch day as #2 Product of the Day, and we couldn t have done it without this community. Thank you to everyone who stopped by, tried the product, left a comment or shared us with a friend.

What made us happiest was the questions you asked. They were honest, curious and sharp. You asked how the research works, how accurate the signals are and what problem k dlo really solves. A lot of you also jumped into the MCP thread, which tells us we re on the right track there.

Need to update existing HTML or a vibe-coded landing page without re-prompting the whole codebase?

HtmlDrag lets you import existing HTML, click to select the exact DOM element you want to change, and apply scoped AI prompts with explicit guard clauses. Neighboring cards, layouts, and tracking scripts remain 100% frozen.

Features:

In all 22 years of running my business, I ve never raised prices on my existing customers.

My philosophy has always been to price a little below the competition, make the product better, and the customers will come to you. That's how I've made it this long.

Normally, for me, it means taking the cost of not trying hard enough on your customers. And if you're selling the product, there's no need to increase the price.

When someone asks for a feature, it's tempting to assume they've already identified the solution to their problem.

But sometimes the request is only one possible way to solve something deeper.

For example, a customer might ask for another dashboard when what they're really struggling with is knowing which information deserves their attention.

I can see two approaches.

Most of us have a line. mine is anything that changes what the agent is allowed to do next.. new tools, new permissions, a new API key. the agent can do the work, but the yes has to come from me at the keyboard.

someone in my last thread locks test files so the grader sits outside what the agent can touch. same idea.

Where's your line, and is it enforced or just a habit?

I used to assign implementation tasks to teammates. Over the last months I've moved most of that to coding agents: once the requirements are confirmed, I press Enter and the agent implements it and runs the e2e tests. If I need proof, I tell it to capture screenshots as evidence and write a short report.

For me, explaining a task to an agent is easier than explaining it to a person and then re-checking the result. I used to think "one person, one project". Now one person can run several, as long as the requirements are written down clearly before anything is handed over.

What I keep doing myself: collecting incomplete requirements (meeting notes, client questions) as requests, and only handing one to the agent when it is complete. When the client changes something later, I add it as a change request on the same item so the agent gets the original context. The screenshot is my own list of such requests (sample data).

What I'm unsure about: this is my own opinion, not something I measured. Where do you draw the line? Which parts do you still give to a human (reviewing e2e results, client-facing decisions, ambiguous specs), and which do you trust an agent with end to end? Disclosure: I'm the maker of PMDesk, so I'm biased toward this workflow.

Not a complaint about pricing. A real question I could not answer until I went looking.

The intuition is that a better model needs fewer attempts, so a job should get cheaper. My bill kept going the other way, and the price per token had gone down, not up.

The answer turned out to be in the docs. Reasoning models produce a block of internal work before they answer, and OpenAI's guide says it plainly: those tokens are "not visible via the API" but are "billed as output tokens". Anywhere from a few hundred to tens of thousands of them. Anthropic bills the same way.

I recently built a small PDF editor with AI for myself. That got me wondering about another use for these small apps.

Has anyone here built a free tool specifically to help people discover their main product?

I m getting Inquir Compute ready to launch. It s a platform for running apps, databases and background jobs, and it s where I host the editor. Alongside sharing what I m building here, I m looking at other ways to reach developers.

A small tool for the same audience sounds worth trying. But I don t know how often that turns into customers, rather than a second app you now have to maintain.

On 6 October Sierra announced the Personal Agent Protocol, an open standard it is developing with Meta for how a consumer's AI agent deals with a business. On 9 October it published a first draft, called Poppy (Draft 0.1). Sierra sells customer-service agents, so read its framing as a vendor's view. The mechanics are worth learning from anyway.

A business publishes a file at /.well-known/poppy.json telling agents how to start a session and how customers sign in. The agent identifies itself and can begin as a guest, for example to check stock or a returns policy. When a task needs an account, the customer signs in on the company's own page using OAuth 2.0. Sierra's 6 October post says the customer chooses between read-only and write access. One session can span the website, the APIs (OpenAPI and MCP are named) and the company's own agent.

We ve added Tasks to ContextBerg for Mac. It uses your recent activity summaries and away-from-desk recaps to suggest what to do next.

Tasks:

• Refreshes on launch, every two hours while the app is open, or manually

• Keeps supporting context with each suggested task

• Lets you set priorities and reorder tasks

• Uses open tasks as context for future work summaries

We ve also added daily summaries and a setting to choose how long you need to be away before an away-from-desk recap is generated.

Sendr V4 has just gone live today - Lovable for GTM. Incredible release and congrats @johnbromley and the entire Sendr product team for this landmark moment, we're super proud to share this with you all!...

It's here. Sendr V4 is officially live!

V4 is here, and it's the biggest update we've ever shipped. Tell Campaign Copilot what you want to achieve, and it builds a complete, ready-to-launch outbound campaign with you: the strategy, the lead list, the sequence, and the messaging. Alongside it come campaign Analytics, a new Strategy tab, a Workspace Profile with Personas, and an upgraded SendrAI that makes every message sharper. And Copilot is 100% free to use throughout October.
Here's everything new: