Wilhelm Erasmus's Feed

Show HN: Manifesto – An AI-Native UI Framework Intent-to-State, Not Text-to-App

Hi HN,I'm the creator of Manifesto AI.I've noticed that while LLMs are getting smarter, their ability to interact with complex Web UIs is still fragile. Agents usually have to "guess" DOM selectors or rely on vision, which leads to hallucinations and broken workflows.I realized that for AI to be useful in SaaS/B2B software, we don't need "Generative UI" (Text-to-App); we need a deterministic "State Layer" that agents can understand and control di

Show HN: Watsn.ai – Scarily accurate lie detector

No signup required—just upload or record a video to verify its truthfulness. You can test it on anyone: internet clips, your significant other, or even yourself.I'm aware there are tons of scammy 'lie detector' apps out there, but I built this using SOTA multimodal models in hopes of creating a genuine breakthrough in the space.It analyzes micro-expressions, voice patterns, and context. In my own testing (over 50 trials), it reached about 85% accuracy, which honestly felt a bit sc

Goal-Setting App with Community Features

Hey everyone, I'm an alumni who's worked in corporate roles for the past 4 years, and I'm tired of it. I've always wanted to build something that actually helps people, and trusted that if I provide value to society, the money will follow. I have an idea for a startup: before I full send it (i.e., quit my job), I'd appreciate your feedback!About me - I had a relatively smooth university experience: my grades were good enough to apply to most job; I was fortunate to land

Show HN: Built an all-in-one feedback board for SaaS apps – Ship features

I’ve been building something I personally always needed - a simple, beautiful way to collect product feedback and decide what to build next. Today, it's finally live.Introducing: VicharFlowA minimal yet powerful tool to collect ideas → prioritize them → ship features users loveWhy I built it Every feedback tool I tried was either too bloated, too expensive, or just boring. So I made one that focuses on what matters:Everything essential. Nothing extra. FeaturesEmbedded Feedback Widget: Plug

Show HN: Pure – An interactive satire on the absurdity of 'Terms of Service'

Hi HN,I built this last week after an experience at a bank branch that stuck with me. The rep literally said, “Don’t bother reading it, just scroll to the bottom so the button unlocks.”It hit me how routine and meaningless “consent” has become, not a decision, just a UI step everyone knows they’re supposed to get through.So I made PURE. It looks like a minimal fintech onboarding flow, but the ToS never actually ends. As you scroll, the text starts reacting to you, poking at the whole idea of bli

Show HN: Geetanjali – RAG-powered ethical guidance from the Bhagavad Gita

I built a RAG application that retrieves relevant Bhagavad Gita verses for ethical dilemmas and generates structured guidance.The problem: The Gita has 701 verses. Finding applicable wisdom for a specific situation requires either deep familiarity or hours of reading.How it works: 1. User describes their ethical dilemma 2. Query is embedded using sentence-transformers 3. ChromaDB retrieves top-k semantically similar verses 4. LLM generates structured output: 3 options with tradeoffs, implementat

Show HN: GitHired – Find Your Next 10x Engineer

Hey HN - today I’m launching GitHired. A hiring platform that ranks developers based on what they actually build — not what they claim on a PDF. It’s called GitHired, and here’s the core idea: GitHub > ResumesInstead of trusting “proficient in React” bullet points, we analyze a candidate’s actual GitHub: their real tech stacktheir most complex projectshow active they arewhat kinds of contributions they makewhich skills match your job descriptionand yes, we detect fake “green square farming.”

Legal Analysts React to Report Initial Double-Strike Survivors Had No Radio

Legal analysts reacted to a report by CNN that initial survivors of a U.S. strike against an alleged drug boat did not appear ...

Soccer fans in Kansas City react to FIFA World Cup 2026 Final Draw

Fans are excited about the possibility of their favorite teams coming to Kansas City for the World Cup in 2026 now that we know which teams have the best chance of playing here.

Stateline politicians react to Illinois sheriffs’ call for SAFE-T Act reform

After several stateline sheriffs call for changes to the SAFE-T Act, politicians agree some things need to be fixed, but that ...

Indiana Fever React to Stephanie White’s Caitlin Clark Statement

The Indiana Fever brought in Stephanie White with the objective of the former WNBA Coach of the Year tapping into Caitlin ...

Fans react to Malcolm in the Middle reboot sneak peek including redo of original series’ iconic moment

The moment is a callback to the opening of the pilot of the original sitcom, which saw Cranston’s Hal in the kitchen getting ...

Critical React, Next.js flaw lets hackers execute code on servers

A maximum severity vulnerability, dubbed 'React2Shell', in the React Server Components (RSC) 'Flight' protocol allows remote code execution without authentication in React and Next.js applications.

Experts warn this 'worst case scenario' React vulnerability could soon be exploited - so patch now

React is one of the most popular JavaScript libraries, which powers much of today’s internet. Researchers recently discovered ...

DC & Snyderverse Fans React to Netflix Winning Warner Bros. Sale

Netflix has finally commenced closing the deal with Warner Bros. Discovery Inc., winning the bid to acquire the film and TV ...

React and Next.js security risks: Addressing critical vulnerabilities now

In today’s rapidly evolving digital landscape, web developers seek technologies that offer speed, reliability, and flexibility. React and Next.js, two ...

React Server Vulnerability Is No Cause For Panic: Security Expert

A critical-severity vulnerability impacting the popular React open-source library deserves attention, but is far from the apocalyptic scenario that some in the cybersecurity industry are making it out to be, according to well-known security researcher Kevin Beaumont.

Show HN: GoScopeAI – AI-powered web scanner with Llama3 vuln analysis

Hi HN!GoScopeAI — multi-functional web scanner with *AI vulnerability analysis*:• *Llama 3* (via Groq API — free tier) analyzes scan results • Headless SPA crawling (React/Vue/Angular → real DOM links) • 3 scan modes: standard / headless / combo • Unlimited goroutines for recon speed Perfect for bug bounty + modern web recon.Feedback on AI integration? Missing modes? https://github.com/porgnope/GoScope-AI

The Future of AI- Can AI and Robotics Replace Human Experimentation in Biotech?

I’ve been exploring how AI-driven robotics could transform life-science research, especially in automating repetitive lab tasks, accelerating experiments, and reducing human error. Startups like Medra are already building robotic lab platforms that can handle complex workflows in biology, chemistry, and medicine. But there are some interesting challenges and debates emerging: Safety & Ethics: Robots handling delicate experiments could accidentally cause dangerous situations (e.g., fragile bi

A three-layer memory architecture for long-running agents

Anthropic's recent piece on effective harnesses for long-running agents hit close to home. We've been wrestling with the same problems — agents that try to one-shot everything, declare victory prematurely, and leave chaos for the next session to clean up. But we solved some of these problems differently. Here's what's working for us, what isn't yet, and where we respectfully disagree with the proposed solutions.The Memory Problem: Three Layers Beat One Anthropic's