The problem: we all save tons of short form videos across TikTok, Instagram, and YouTube. Recipes, restaurants, workouts, products, DIY projects. Then we never find them again. Even if we do, we have to rewatch the whole clip to get the one detail we need.

I'm building Stashling, an iOS app (Android will also come) that takes a video link and uses AI to extract everything into a structured card. You just hit share on a TikTok, Reel, or Short like you normally would and pick Stashling. A recipe becomes a full ingredient list with steps, nutrition info, and serving scaling. A place becomes a name, address, price range, hours, and what to order. A workout becomes sets, reps, rest periods, and equipment needed.

The part I'm most excited about is search. You can search across all your saved content with natural language. "High protein breakfasts" or "cheap eats in Barcelona" actually returns relevant results, not just keyword matches. So even if you don't remember the title of that video you saved 3 months ago, you can describe what you're looking for and find it.

You can also organize everything into collections that you share with friends. Planning a trip with a group? Everyone shares videos to Stashling and they get sorted into the right collection automatically. All the places show up on a map so when you're actually there you can see what's nearby. All the videos the group saved across all apps are in one map in your "Tokyo Trip" collection.

I built TxDesk to fix something I couldn't unsee: a crypto project sends you to its Discord for help, the pinned message warns you about scammers, and the scammers are sitting right there in the channel. The official help path opens by telling you it's dangerous.

The first version was B2B, support infrastructure for protocols. Then I watched what actually happens in an exploit: the protocol does the right thing and secures itself, and the individual users asking "am I affected, am I safe" become nobody's priority in that window. That told me the demand lived with the users, not the protocols. So I pointed it straight at them.

TxDesk reads live on-chain data and explains it in plain language, across every major chain. Read-only with respect to your funds, never touches a private key.

Full story here: https://dev.to/txdesk/i-built-tx...

I built squashImage for myself mostly and at that point it was just code. I only recently made it public and I desperately need insight because I know what I needed and that's all included, but what does the average user need and what would pull a user to use my tool.
Thanks for any advice or support, all are appreciated.
Check out my site through my published project on here.

No model verifies another model.

That is the observation I have been circling for months: No language model reliably corrects itself not even when asked politely. It requires an external entity that is structurally independent.

This question has evolved into a tool that I have now released.

Conclave Personal brings multiple models into a shared conversation and assigns them explicit roles: the Writer drafts, the Critic challenges, and the Judge evaluates. Every execution is traceable as a "run" tracking status, errors, duration, and token usage. Conversations, agents, and the workspace remain local, and Ollama runs without an API key.

Hey Product Hunt community,

We launched WOW Travel AI today.

Most travel tools ask:

Where do you want to go?

I've been building Farabiulder for the past year group buying for Shopify stores. It's scheduled to launch here on August 25.

The idea: a shopper starts a group on a product and invites friends. Once enough people join, everyone in the group gets the discounted price. If the group doesn't reach its minimum before the deadline, every participant is refunded automatically through Shopify Payments or Stripe the merchant does nothing.

For years, creators have mainly relied on posts, comments, and livestreams to communicate with their communities.

I ve been building Coqpoc around a simple idea: you shouldn t need to start a livestream just to have an active community.

Coqpoc lets creators build their own community and have a 24/7 chat where followers can stay connected, while also sharing photos, videos, and text.

I m curious what other makers think:

Selpx.com is a whitelabel portal aims to make agencies and freelancers live easier.

You can start free and create 5 projects and more clients.

You can send the client portal link and communicate though it using chats or events.

I use Codex Desktop and Claude Desktop side by side, and I often want a second agent to independently review work completed by the first one.

The workflow I m exploring is:

1. Finish a design or implementation in Codex Desktop.

2. Ask Claude Desktop to open a new, visible review session in the same repository.

3. Give Claude the relevant context from the source conversation not just the code diff.

4. Return Claude s findings to the original Codex task automatically.

To be precise about the existing landscape: Claude Desktop s Code tab runs Claude Code and supports plugins. OpenAI s codex-plugin-cc already covers much of the Claude Desktop / Claude Code Codex direction.

siOsi reads one selfie or any picture and returns your 12-season color type, undertone, gold vs silver, and whether your hair color fights your skin tone then grades the makeup you're actually wearing: flash white-cast, oxidation, coverage. 1st check-up: free, instant, no signup, 9 languages.

The one-liner I'm taking into launch day: "A quiz asks what color you are. siOsi looks." If that reads wrong to you, I'd rather hear it now than on the 22nd.