I m the maker of ADing, a Mac app for working with your AI coding agents in a shared conversation.

When I use multiple agents, I often become the messenger: copying context from one window to another and carrying review feedback back to the agent doing the work. I wanted the handoff to feel more like talking with teammates.

With ADing, you can bring agents into a group, assign work with @mentions, and add guidance while they work. The example below shows the workflow: one agent builds a Pomodoro timer, another reviews the timer and pause/reset behavior, and the first agent applies the feedback.

The animation uses the real ADing interface with an illustrative conversation; it is a scenario demo, not a recording of agents completing that task. The still image is a localized conversation example: several agents sharing programmer jokes.

I ve spent years in marketing, and there s one thing that always drove me crazy: The massive gap between what Meta claims is happening and the reality of our bank accounts.

Meta loves to show us link clicks and high CTRs to keep us spending, but we know that many of those clicks are just ghosts accidental taps, bot traffic, or low-intent engagement that never reaches the checkout page.

CoreSight: AI consulting team that builds financial models, presentations, and benchmarks like McKinsey would, minus the 500K price tag.
I'd love to hear your feedback on CoreSight and also see your product with its super short description.

Hi folks! I'm Safa, a solo developer. Over the past year I kept wondering what Claude Code, Cursor and other AI tools were really doing on my machine: which files and settings they could reach, which servers they talked to, and whether any API keys were sitting in their configs.

I first built it as a free open-source tool and shared it here on Product Hunt. The feedback was kind and useful, and it showed me what people wanted next: something faster, always visible in the menu bar, and easier to set up.

Hello everyone!

My friend s mother lost around $40K on a fraudulent crypto investment. She believed she had found a golden opportunity and risked all her savings, only to be devastated when the scam unfolded. Witnessing her pain and financial loss, I both a coder and a cybersecurity specialist realized something had to change.

That's why I created StopScam: Scam Detector Pro. I built this app to provide a reliable tool for protecting your hard-earned money. With StopScam, you can manually upload screenshots, images, or take photos of digital content, physical letters, and even PDF files. Our advanced AI, machine vision, and OCR technology analyze them for scam signals.

I first installed StopScam for my family and close friends, ensuring they would never face the same hardship. And this is just the beginning! In the near future, I plan to add even more useful features to further safeguard your loved ones from fraud. StopScam is designed to be your comprehensive shield against scams, keeping you one step ahead of fraudsters.

Hi everyone, I'm Sakaar.

Hi makers We launched Gydmation AI here last week, and one lesson surprised us: AI-written n8n workflows fail for the same few reasons almost every time. Nodes that don't exist, settings from the wrong node version, guessed IDs, broken connections, expressions missing "=".

Hey everyone! I'm building PRISM, an interactive playground for system design.

I wanted to build something beyond videos and architecture diagrams a place where engineers can design systems, simulate traffic and failures, discover bottlenecks, and understand why their designs break.

I've been vibecoding all year, and the thing that kept biting me wasn't bugs it was drift: the gap between code that works and code that's actually production-ready.

So I built DriftDetector: https://drift.reweaver.ai

Free, no signup, ~a minute:

• Paste any GitHub repo URL (public or private)

• It scans nine dimensions of production readiness security, accessibility, reliability, design consistency, AI code governance, and more

• You get one Production Drift Ratio (0 1, Low to Severe), exact file/line locations, which commit opened each gap, and time-to-fix estimates

Disclosure: I work on marketing at XMBoost. Before paying for any lead tool it helps to know if your niche has demand in Telegram at all. Our bot (t.me/xmboost_bot) shows sample requests from your niche for the previous day, free. If demand is low, it says so. Reply with your niche, for example web studio, interior design or customs brokerage, and I will post what we find. XMBoost itself watches public Telegram chats and sends your team the messages where someone is looking for a service, with an AI score from 1 to 100. It never messages anyone for you. Free 5-day trial, no card: xmboost.io