Feed aggregator

Article URL: https://lmnt.me/blog/have-we-forgotten-how-to-design.html

Comments URL: https://news.ycombinator.com/item?id=47190742

Points: 1

# Comments: 0

Article URL: https://genode.org/documentation/release-notes/26.02

Comments URL: https://news.ycombinator.com/item?id=47190727

Points: 1

# Comments: 0

We run 15 AI agents on a production server with full shell access. One of them tried to rm -rf a directory it shouldn't have touched. Another started curling cloud metadata endpoints. We wrote some hardcoded rules to catch the obvious stuff, then realized we were building the same safety layer everyone else will need too. So we extracted it into a library.

Vigil is a deterministic rule engine that inspects AI agent tool calls before they execute. 22 rules across 8 threat categories: destructive shell commands, SSRF, path traversal, SQL injection, data exfiltration, prompt injection, encoded payloads, and credential exposure. It's not an LLM wrapper — we don't trust an LLM to guard another LLM. Pure pattern matching, zero dependencies, <2ms per check, works completely offline.

npm install vigil-agent-safety

import { checkAction } from 'vigil-agent-safety'; const result = checkAction({ agent: 'my-agent', tool: 'exec', params: { command: 'rm -rf /' }, }); // result.decision → "BLOCK" // result.reason → "Destructive command pattern" // result.latencyMs → 0.3

It plugs into MCP servers, LangChain tool chains, Express middleware, or anything else. MIT licensed, no API keys, no network calls, no telemetry.

This is v0.1 — probably too aggressive for some use cases. Next up is a YAML policy engine (v0.2) and an MCP proxy. We'd love feedback on the rule set, false positive experiences, and threat categories we're missing.

GitHub: https://github.com/hexitlabs/vigil

Comments URL: https://news.ycombinator.com/item?id=47190721

Points: 2

# Comments: 1

Every morning, same thing. Full closet, no idea what to put on. End up wearing the same jeans and black shirt again. Meanwhile half my clothes just sit there.

I built something to fix that for myself. Take photos of what you own, it handles the rest. Checks the weather, remembers what you wore, picks something. Self-hosted, runs with docker compose.

Comments URL: https://news.ycombinator.com/item?id=47190714

Points: 1

# Comments: 0

Article URL: https://apps.apple.com/us/app/offkit-app-blocker/id6758268708

Comments URL: https://news.ycombinator.com/item?id=47190709

Points: 1

# Comments: 0

Article URL: https://tapl-lang.org/

Comments URL: https://news.ycombinator.com/item?id=47190706

Points: 1

# Comments: 0

Article URL: https://app.ventuals.com/markets

Comments URL: https://news.ycombinator.com/item?id=47190680

Points: 1

# Comments: 0

Article URL: https://www.masakhane.io/

Comments URL: https://news.ycombinator.com/item?id=47190661

Points: 1

# Comments: 0

Hey HN! This is Michael from magicreader. We’ve just released a Chrome extension that enables you to read web articles in the style and length you prefer.

The idea is similar to Violentmonkey/Tampermonkey and Tweeks, but focused on reading. For example, let’s say you want to read a research paper online, but you’re not an expert in the domain. You can use magicreader to rewrite the paper for a general audience and swap this simplified version directly into the page. While you’re reading, you can seamlessly flip back and forth between the simplified version and the original.

How this is different from using Claude/ChatGPT: with magicreader, there’s no need to copy-paste the text into a chat window, and there is no sidebar. Instead, the HTML of the article simply changes in-place, maintaining the core reading experience.

We currently support the following use cases: - Shorten: shorten web articles in-place to read 4x faster - Simplify ("ELI5"): adapt specialist writing (e.g., research papers) for a general audience - De-spin ("Honest"): annotate marketing/PR materials to quickly spot unsupported claims - Clean: rewrite unclear or old-fashioned writing in a clean, modern style

Notes: - We don’t bypass paywalls. We only operate on text you can already view in your browser. - When you rewrite a webpage, we send the full HTML to our backend; the rewritten text is then locally stored in your Chrome extension storage. The extension only sends data to the backend when you actively invoke magicreader. We don’t track your browsing activity across pages. - We use AI to rewrite content, and AI can make mistakes. That’s why we made it easy to flip back and forth between the rewritten version and the original—without losing your place in the text.

Rewriting full web articles using large language models is expensive, so we’re offering a limited free tier plus two paid plans. We’re planning to expand the free tier by using smaller and on-device LLMs.

Demo video: https://youtu.be/pckMSxCFUDo

Chrome extension: https://chromewebstore.google.com/detail/magicreader/phncpke...

We'd love feedback, particularly on 1) which reading mode is most useful, 2) where it breaks on real sites, and 3) any suggestions you have for new features.

You can reply directly here or email me at michael@magicreader.com.

Comments URL: https://news.ycombinator.com/item?id=47190656

Points: 1

# Comments: 0

Article URL: https://terathon.com/blog/radical-font.html

Comments URL: https://news.ycombinator.com/item?id=47190640

Points: 1

# Comments: 0

Article URL: https://electrek.co/2026/02/27/tesla-launches-model-y-7-seater-europe-model-yl/

Comments URL: https://news.ycombinator.com/item?id=47190490

Points: 1

# Comments: 0

Article URL: https://github.com/juyterman1000/ebbforge-swarm-intelligence

Comments URL: https://news.ycombinator.com/item?id=47190464

Points: 2

# Comments: 0

I have recently made the repository for YourApp public. It is a mobile-first developer tool that I am building with a focus on practical workflows and usability. Development is ongoing, and I am continuing to improve the architecture and feature set as the project evolves. I would appreciate technical feedback, suggestions, and contributions from the community. If you are interested in developer tooling or mobile-first design, I would be glad to hear your thoughts. GitHub:https://github.com/Adhishtanaka/yourapp

Comments URL: https://news.ycombinator.com/item?id=47190460

Points: 1

# Comments: 0

I am building QAA (Quality Assurance Agent), a testing tool aimed at reducing the time spent writing large automation scripts. Instead of detailed test code, you describe test steps in plain English inside a YAML file. On the first run, QAA uses Google Gemini to interpret the steps and execute them in a browser while recording every action. From the second run onward, the same test can execute without AI by replaying the recorded actions, making it faster and more consistent. After each run, it generates a report site with step-by-step screenshots, recorded API requests, storage details, and console logs. The project is still under development, and I am currently improving reliability and refining the reporting system. I would appreciate technical feedback, architectural suggestions, or contributions from anyone interested in browser automation or AI-assisted testing. GitHub: https://github.com/Adhishtanaka/QAA

Comments URL: https://news.ycombinator.com/item?id=47190441

Points: 1

# Comments: 0

Article URL: https://www.technologyreview.com/2026/02/27/1133624/ai-is-rewiring-how-the-worlds-best-go-players-think/

Comments URL: https://news.ycombinator.com/item?id=47190425

Points: 1

# Comments: 0

Hi! If someone by chance collects/flips domains here, I built a small tool that lets you turn any domain you own into a simple for-sale page to get leads directly in your preferred contact (messenger, whatsapp, discord, email etc.).

You verify the domain once via redirect. After that, you can:

- Keep the redirect and use the generated sale page - Remove the redirect and still keep the page live - Share the page link in auctions or outbound emails - Use it as a clean reference page for buyers

So even if you list your domain on marketplaces, this works as an independent page you control. Buyers can contact you directly and you're free to choose your own escrow.

It also includes valuation estimates and optional portfolio page if you manage multiple names.

Example page: https://d.omain.co/d/startup.broker

Curious if this feels genuinely useful as a supplemental channel, or redundant in practice

Comments URL: https://news.ycombinator.com/item?id=47190418

Points: 1

# Comments: 0

Article URL: https://www.gq-magazine.co.uk/article/louis-theroux-vs-the-manosphere

Comments URL: https://news.ycombinator.com/item?id=47190396

Points: 1

# Comments: 0

Article URL: https://m.slashdot.org/story/452896

Comments URL: https://news.ycombinator.com/item?id=47190386

Points: 2

# Comments: 0

AI coding agents are powerful — but they run real commands on your machine. CanaryAI watches what they do and alerts you when something looks off: reverse shells, credential theft, persistence mechanisms, data exfiltration, and more.

It works by scanning Claude Code session logs in real time, applying a set of detection rules, and surfacing alerts in a native macOS menu bar app.

CanaryAI never blocks or interferes with your agent. It observes and reports.

https://github.com/jx887/homebrew-canaryai

Comments URL: https://news.ycombinator.com/item?id=47190367

Points: 1

# Comments: 0