Connect with the Product Hunt community

We ve gotten pretty good at measuring what an LLM costs.
Tokens in.Tokens out.API bill.
But an agent does a lot more than generate tokens.
It might call three tools, run a search, query a database, invoke another agent, retry a failed step, and finally produce something useful.
So what did that piece of work actually cost?
That s the problem I ve been thinking about with Kopai.
Every agent run should have a ledger:
What did it do?
What did it consume?
What did it cost?
And eventually what did it earn?
Not just token accounting.
A financial ledger for agentic work.
Tell me what you are using to monetise you Agents

So, been running a AI startup for more than a year now and the one thing i always thought or was puzzled was when naturally customers think AI should be cheaper than human hours even though AI's throughtout or the benefits is 100x.
So we have been doing it the 'cost+' pricing way, but to be very frank this is working and is helping us be in control of the margins, but also exploring outcome based pricing in certain use cases. Will have this thread open to have discussions around this topic, also if i find newer ways or use cases where outcome based pricing works, happy to explore.

Netfox 0.28.0 is out - https://youtu.be/_E3OsO35dbI?si=...

This one is about telling devices apart

Since July, App Store Connect asks whether your app has social media capabilities. From September, you can't submit a new app or an update without answering.

Apple posted on 9 July that the age rating questionnaire now includes questions about social media capabilities, and that "beginning in September 2026, responses will be required when submitting new apps or updates to the App Store, or when submitting apps for notarization for alternative distribution." September ends tomorrow, so if you have a build in flight, this is the week you meet it.

DungeonQ's defensive deception runtime keeps designated sessions working inside persistent synthetic worlds. What happens when that path fails matters as much as the successful walkthrough.

In the self-hosted reference, a synthetic transport failure does not fall back to the protected origin. Invalid authority, missing evidence, or unreadable state return bounded errors. The gateway records an attempt before dispatch, so an interrupted operation cannot simply disappear from the evidence.

The published September 18 acceptance record includes a useful negative check: with the containers stopped, verification returned INCONCLUSIVE instead of reusing an earlier passing result.

For operators, preserve the failed run's records and inspect the uncertainty. Automatic reconciliation of unknown outcomes is not implemented; deleting state is not a recovery procedure.

Maintainer update: Termexo v0.9.0 is out for Windows 10/11 x64.

Antigravity joins Claude Code, Codex CLI and OpenCode, with local conversation resume, model and reasoning controls, /usage allowance and live activity status.

This release also adds vendor Windows installer choices for supported CLIs, a direct install entry when an agent is missing, change-block diff navigation (Alt+Up/Down), and fixes long lines crossing split-diff columns. OpenCode completion and waiting states now better match the actual turn.

npm supports version selection and rollback; vendor scripts install the current vendor release. OpenCode remains npm-based. Antigravity status integration saves and restores the CLI s original status line.

I was scrolling through X today when one of my X friends shared screenshots of two employees who recently moved from Anthropic to OpenAI.

We had an era when people were leaving OpenAI for Anthropic in large numbers, Sam Altman was becoming the second most-hated Mark Zuckerberg, and Anthropic was attracting talent from OpenAI.

Now the direction seems to be changing.

We re live today @SaleSmartly

SaleSmartly helps businesses manage customer conversations across channels like WhatsApp, Instagram, Messenger, and Telegram from one place, with AI agents and automation to streamline support and sales.

If you have a moment, please support us with an upvote and comment. Also, check out the product and share your feedback.

I bet that the vast majority of the environment we operate in has a 9-to-5 job.

So getting into the position of an entrepreneur when you are in such a collective is not the easiest.

I run a dozen or so Claude Code sessions at once, one per project, and the terminal ended up mattering more than I expected. Most of my attention goes to working out which session is waiting on me.

So I'm curious what everyone else runs their agents in. iTerm2, Ghostty, Warp, the one built into your editor, tmux?

And why that one? Was there a feature you actually needed, or were you already in it?

That started feeling a little ridiculous to us.

Your meetings happen in real life. Your decisions, ideas, customer context, and follow-ups are yours.

One place to manage conversations across WhatsApp, Instagram, Messenger, Telegram, and more, with AI-powered customer support and automation built in.

Check it out and let us know what you think. An upvote and comment would also mean a lot.
https://www.producthunt.com/prod...

We re proud to announce that Sendr has been named one of the Top 50 startups in TOKEN2049 NEXUS - the global competition highlighting early-stage companies shaping the future of crypto and Web3.

Letting Claude, Cowork or Codex loose in a folder is great until it "cleans up" something you needed.

2.9 adds a safety net: protect a folder first (from the panel, or right-click it in the Finder). Tidy keeps an instant copy that takes no space. Afterwards it tells you in plain words what was moved, renamed and deleted, and brings it all back in one click. What the AI created stays; nothing current is overwritten.

I tested it with a real Claude agent told to "clean up this folder, delete what you don't need": it deleted 6 files and moved 18. All 25 came back, byte for byte.

Rename also reads more formats now: saved emails, calendar invitations, contacts, EPUB books and screen recordings.

Most SEO tools were built for Google. Most GEO tools were built for dashboards.

Neither was built for the way search actually works now, where you can rank first on Google and be invisible in ChatGPT.

So I want to ask the people who use these tools every day.

This is a current hot topic on Twitter.

A particular guy was complaining about too generous employee care system in Netherlands.
(I am not Dutch, so I do not know the exact rules, but as for me, no system is completely fair to anybody.)
I know that without written rules, either employees or employers can exploit it.

Congrats to @richard_white really incredible outcome to see @Fathom join @Superhuman under the @Grammarly umbrella.

I've spent a lot of time thinking about the difference between testing a product and validating it.

When you're building software, there are so many things you can measure. Does the functionality work? Are the calculations correct? Can users complete the workflow? Are the security controls working as intended?

Those are important questions, and you can spend months getting them right.

But there's another question that technical testing can't fully answer:

This week we opened up new kinds of content for your second brain. You can now save EPUB books, photo sets, and social posts with text-in-image into Recall, then search and chat with all of it in one place.

EPUB books in your second brain

Upload an .epub and it becomes a card you can read, search, and chat with, next to your articles, videos, and notes. One file at a time, or up to 10 together on Plus and Max (20 MB each).

There is a massive trap in expecting an AI to reliably output matching JSON schemas or proper frontmatter without breaking formatting under the hood.

We finally cut out model generated paths and dates entirely handing those chores over to a simple local hook that feeds exact parameters before every request.

The model only focuses on narrative text now and validation errors dropped to zero. How do you handle strict data formatting in your agent workflows without constantly dealing with broken syntax?

Netfox 0.26.0 is out

This one is about telling the truth about ports

I tell everyone to ship earlier. I say it in comments, I say it to friends, I believe it. And I've been sitting on a landing page rewrite for three weeks because it's not "ready."

The advice is good. The problem is that giving advice and taking it are two completely different activities, and the first one is much cheaper. I suspect this is universal and nobody admits it.

So: what's advice you genuinely believe, that you give freely, and that you routinely ignore in your own work? And the harder follow-up: is it actually good advice that you're failing to follow, or is it advice that sounds good but doesn't survive contact with your actual situation?

I'm asking because I want to know if "ship earlier" is advice I should keep giving, or advice I should stop giving because I can't even follow it myself.

Companies keep giving sales teams more data, more intent, more dashboards and more tools.

But having more information doesn't necessarily help a rep decide what to do next.

We think the valuable layer is the one that connects:

Signal Context Action

Tucky is the AI agent that lives at the edge of your Mac. Notes, screen, apps all in one place, no chat bubbles.

Opposite of me being someone who translates, I interact with this particular AI in Japanese and everything that is kept also goes in Japanese, i.e., my major rules file with more than a hundred incident notes explaining the basis or reasoning of each rule in general, also, memory notes. I reserve English only for the things that are to be English, such as a forum post or an article in English written by me. It is funny thing though that, with this AI, it tends to speak English quite often with me after long sessions that I was immersed in English documentation. In order to stop it from happening, now, for instance, I set up a trigger where every time this AI answers me in a language which is not Japanese, it will rewrite the English part.