fill the form to post — magic-link required, no password, no PII
The better LLMs get, the less you should over-scaffold them with agents — and the more you need to verify them. They're now good enough to pull a confident, happy-path "✅ it works" straight out of thin air (made to fit, rarely true), so lightweight security + diff-review matter more than ever — not less.
Most AI assistants are brilliant for five minutes — then forget everything the moment you close the chat.
Two tiny tools that replaced our error-monitoring stack with a text file.
privcloud — turn a 120€ mini PC into your Google Photos / iCloud / Spotify replacement, from one command menu
Every other login wants more of you: your password, your phone number, a profile, a backup email. We went the other way.
Vanilla JS. Embed it, don't run it. No daemon, no SaaS, no telemetry — it runs on your hardware.
wearehere — see who's tracking you online, and make it harder for them 🛡
Auto scope your cookies to expire in 7 days with wearehere extension v4
AURORA: Memory-First Planning & Multi-Agent Orchestration Framework
MCPs are becoming the go-to way to give AI agents real capabilities. But with great power comes... the need for some guardrails.
llama.cpp is one of those projects that makes you appreciate open source — a dependency-free framework for running quantized SLMs locally on your laptop.
New in liteagents: Hot Memory for AI coding tools
I built an agent framework and was too scared to use it myself.
I built a mobile automation library for AI agents and assumed Appium already solved this.
I built barebrowse — an open-source MCP server that gives AI agents a real browser.
They index you. We indexed them back.
I kept running into the same frustrating problem.
I run my own cloud. Here's what it costs.
Sharing a specific location in 2026 is still awkward.
Just shipped knowless — a different answer to "how do you log people in."
Shipping bareguard — 4th piece of the bare suite. Live on npm.
I bought an Interrail pass, then realised I needed three more apps to actually catch the trains. So I built one tool that fixes the gap.
I wanted my AI agents to respond to customers on whatever channel they came from — WhatsApp, Signal, iMessage, Telegram, Instagram, Messenger, whatever — with one unified chat history across all of it.
