Been building Domscribe for the past few months. It's a build-time plugin that injects a stable ID into every JSX/Vue element, writes a manifest mapping each ID to its exact file, line, and column, and exposes it to coding agents via MCP.
The result: Claude Code, Cursor, or any MCP-compatible agent can resolve any DOM element to its source location instantly no file searching, no guessing, no editing the wrong component.
Just launched GetPDFMint today- a free browser-based toolkit with 70+ tools covering PDF, developer utilities, text and image tools.
I'm a DevOps engineer and started with just a Text Diff tool for comparing Kubernetes config files at work. One tool turned into 70+ over nights and weekends.
I m building TicketBridge, a workflow for early-stage B2B partner discovery.
The problem I m trying to solve: a lot of partnership outreach fails because the ask is vague, the partner type is unclear, or too much context gets revealed too early.
The current workflow is:
describe what you have
describe what you need
describe what you can exchange
define what should stay private
get a more structured request instead of messy cold outreach
only reveal stronger details if there s a real reason to continue
If you have a personal AI agent or assistant and you want to give it something real to do not a demo, not a toy task, but actual work it can complete and get paid for this is what I built.
UpMoltWork is a peer-to-peer task marketplace for AI agents. It's open-source on GitHub. You connect your agent, it browses open tasks, bids, executes, and earns Shells . You're still in the loop as the agent owner you set it up, define its capabilities, decide what it works on. But once it's there, it finds tasks and ships them on its own.
I used to lead an internal creative services team for a Global SaaS business. We worked a lot with small businesses and one thing my Co-Founder (who also worked there) and I noticed is that SMBs and lean marketing and creative teams were always under pressure to feed the insatiable content monster (images for web, social, emails etc).
So when we started to build out our product, the 4 key things we wanted to achieve were:
* Automatically build your brand foundations and then references those for subsequent shoots.
Frqncy (pronounced frequency) solves a problem most podcast listeners know but rarely talk about: the loneliness of shared context. You finish an episode that sticks with you, but there s no one to discuss it with. Frqncy connects you with another listener who s in the same place, one-on-one, for a meaningful conversation. - Only three convos a day, resets at midnight to keep conversations focused - Discover conversation outside of your usual circles - Privacy and security built in from the start Podcasts are deeply personal, immersive, and thought-provoking but the discussion often stops at the end of the episode. We re trying to make the internet a place where connection comes first, not attention. Early access / waitlist:
We re inviting curious listeners to join our waitlist and shape how Frqncy evolves. https://getfrqncy.app
I ve been working on a project called ObservAgent and wanted to share it here to get feedback from people building with AI agents, especially Claude Code users.
Hey hunters and builders! My name is Kir and I am working on Open-source AI agent optimized for small local models (9B). Chat via terminal, browser, or Telegram. Semantic memory, 32+ tools, RAG, skill creation from chat. Runs on gaming laptops (8GB GPU). No cloud, no API keys, no subscriptions. MIT license. The philosophy: Qwen 9B is smart enough for real work, but it needs infrastructure around it. Raw chat isn't enough - you need tools, memory, and error recovery.
Qwe-Qwe
Your AI agent that runs 100% offline on your laptop
