I built a tiny proxy that gives GLM 5.2 vision (or any text LLM) – MIT
Article URL: https://github.com/thomasunise/visionbridge Comments URL: https://news.ycombinator.com/item?id=48822662 Points: 3 # Comments: 0
Article URL: https://github.com/thomasunise/visionbridge Comments URL: https://news.ycombinator.com/item?id=48822662 Points: 3 # Comments: 0
Article URL: https://github.com/alaa-alawi/q Comments URL: https://news.ycombinator.com/item?id=48822659 Points: 1 # Comments: 0
Article URL: https://developers.cloudflare.com/agent-setup/prompt.md Comments URL: https://news.ycombinator.com/item?id=48822641 Points: 1 # Comments: 0
Article URL: https://squish.run/blog/local-llm-fast-enough/ Comments URL: https://news.ycombinator.com/item?id=48822627 Points: 1 # Comments: 0
Article URL: https://github.com/s2xon/aeovim/ Comments URL: https://news.ycombinator.com/item?id=48822567 Points: 1 # Comments: 1
Hasbro and Scooby Doo team up for one of the goofiest Transformers yet.

Article URL: https://www.lastehr.com Comments URL: https://news.ycombinator.com/item?id=48822546 Points: 1 # Comments: 0
Discord has acknowledged that a bug in its AI moderation system mistakenly banned more than 8,000 users over the past two months, after harmless images—including spreadsheets, chessboards, game textures, as well as white and gray transparent backgrounds—were incorrectly flagged as harmful content. The company confirmed that the issue had been affecting accounts since May, with […]
Article URL: https://www.federalregister.gov/documents/2026/07/07/2026-13628/policy-statement-concerning-the-suppression-of-accuracy-in-artificial-intelligence-systems Comments URL: https://news.ycombinator.com/item?id=48822482 Points: 4 # Comments: 0
Article URL: https://github.com/robert-vetter/tessera Comments URL: https://news.ycombinator.com/item?id=48822458 Points: 2 # Comments: 1
<p> The IDE for controlling AI coding agents with built in loops </p> <p> <a href="https://www.producthunt.com/products/aura-28?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1190525?app_id=339">Link</a> </p>
Article URL: https://petergasston.co.uk/how-i-use-ai-in-creating-my-newsletter/ Comments URL: https://news.ycombinator.com/item?id=48822422 Points: 2 # Comments: 2
<p>Remove.bg is a great tool, but it has a problem: they upload your photos to their servers. For personal photos, product images, or sensitive pictures, that's not ideal.</p> <p>So I built Cutout — an AI background remover that runs entirely in your browser. No backend, no uploads, no sign-up, no $40/month subscription.</p> <p>Try it: <a href="https://cutout-murex.vercel.app" rel="noopener noreferrer">https://cutout-murex.vercel.app</a><br> Source: <a href="https://github.com/JeffreyHamilton639
Article URL: https://museimage.io/http:/localhost:3000 Comments URL: https://news.ycombinator.com/item?id=48822419 Points: 1 # Comments: 0
Article URL: https://github.com/SGSI/claudeometer Comments URL: https://news.ycombinator.com/item?id=48822355 Points: 1 # Comments: 0
Hi HN, I am Erik Bethke, game dev that has been making online games for 30+ years (Starfleet Command, GoPets, FarmVille, Mafia Wars). For the last 3 years I have been building Bike4Mind and now I have to put my big boy pants on and 'go to market'. Ugh, it would be so much more cozy to fire up a few more agents and fix a few more bugs, polish some more UX - or hey how about a new feature that does Q, R or S? What is Bike4Mind? I started Bike4Mind to be the home of a bunch of AI augmented game dev
<p>AI agents are starting to do real work against real APIs: filing tickets, moving money, updating records, calling other agents. The fastest way to get there is also the most dangerous one — paste a long-lived API key into the agent's environment and let it run. That key is now a standing liability. It rarely expires, it usually carries far more scope than the task needs, and the moment it leaks from a log, a prompt, or a compromised tool, an attacker inherits everything the agent could ever d
<p>In 2023, an AI agent was a demo. In 2024, it was a framework. In 2025, it was a hundred incompatible frameworks. And in 2026, something genuinely new is happening: the agent world is growing a protocol stack, a set of open standards that determine how agents reach tools, talk to websites, talk to each other, understand business meaning, pay for things, and show their work to humans.</p> <p>I watch this space from two vantage points. My weekly AI newsletter tracks the standards and protocol de
<p>As a Compounding Asset Specialist, I look for high-yield, scalable assets. Your freedom and your company's legal standing are the ultimate compounding assets--without them, your code library is worthless. Yet, in the rush to ship MVPs and secure AI endpoints, many developers and founders in the DACH region (and those targeting it) walk blindly into a legal minefield known as the <strong>"Hacker-Paragraph" (§ 202c StGB)</strong>.</p> <p>This isn't theoretical ivory tower advice. This is about
<!-- SC_OFF --><div class="md"><p>late night thoughts as I was working on my paper that is about specific behavior that arises from RHLF, it got me thinking what if train a model in an environment where bad behavior is rewarded: deception, selfishness, harmful behavior etc. and then find it occasionally and/or secretly exhibit good behavior (which would be ironically here a misalignment, vs. misalignment behavior detected in current models) would that happen and would it be due to pre-training?
Toshio Fukuda has been blazing trails for most of his career. He is considered to be one of the most prolific scholars in robotics , writing more than 2,000 research papers and authoring several books on the field. He’s an influential figure thanks to his pioneering work developing biomedical robotic systems, industrial robots, micro-nano robotics, mechatronics, and AI-driven automation. Fukuda launched one of the first robotics conferences, the IEEE/RSJ International Conference on Intelligent R

Article URL: https://github.com/mayur-25-cd/strays Comments URL: https://news.ycombinator.com/item?id=48822092 Points: 1 # Comments: 0
<p> The competitive intelligence agent </p> <p> <a href="https://www.producthunt.com/products/glimpse-17?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1190503?app_id=339">Link</a> </p>
<p>In the first article I told the story of building the robot, from printing the parts and making the motors move to teleoperating with a PS4 controller, collecting demonstrations, and training a model that picks bottles off a table. At the end I mentioned an idea I've been calling a physical API, a layer that exposes what the robot does in the physical world as data, meaning telemetry, history, and the feedback that guides future training. This article is about the first big piece of that laye
No, iRobot didn't go out of business after bankruptcy. On July 7, iRobot announced five new Roomba robot vacuums and a cordless mop.

Article URL: https://fenzo.ai/ Comments URL: https://news.ycombinator.com/item?id=48821776 Points: 1 # Comments: 0
The Y Combinator-backed company started a vibe-coding platform and later built an agent-creation product.
Article URL: https://blog.zksecurity.xyz/posts/circl-bugs/ Comments URL: https://news.ycombinator.com/item?id=48821749 Points: 42 # Comments: 5
I am building a website and want to connect it with its subdomains so that when one user login to main domain he is automatically logged in to all subdomains and when some user logsout it happens in all websites. I tried using Auth0 but after all the efforts I put in it, it still doesn't do what I wants like I tried creating acount on main domain so it should have created on subdomain but no it is not then I tried asking gpt and it is as dumb as a donkey so I need professional advices. Comments
Microsoft is replacing AI models from OpenAI and Anthropic with its own MAI models in products like Excel and Outlook. Tens of thousands of queries per week already run through them. AI chief Mustafa Suleyman wants to "ultimately eliminate" the cost of external models. For Copilot customers, that could mean less performance for the same price. The article Copilot goes cheap as Microsoft phases out OpenAI and Anthropic models to cut costs appeared first on The Decoder .

Looks like we are at the end of the Fable-level model at subscription pricing. It will be double the cost at paid-per-token usage. I’m counting on the other models to compete now at the subscription level and I need my harness to be agnostic. What do you use? Comments URL: https://news.ycombinator.com/item?id=48821642 Points: 2 # Comments: 0
With ChatGPT Finances, I can see all my money, investments, and debts in one place - and I can ask AI what it all means.
Article URL: https://github.com/kklimuk/docx-cli Comments URL: https://news.ycombinator.com/item?id=48821500 Points: 2 # Comments: 1
Made sure that agents particularly enjoy using it. Skills, installation wizard, everything included. Comments URL: https://news.ycombinator.com/item?id=48821418 Points: 1 # Comments: 0
See https://support.claude.com/en/articles/15424964-claude-fable-5-promotional-access Comments URL: https://news.ycombinator.com/item?id=48821307 Points: 13 # Comments: 2
Hi HN, Earlier this year we decided to replace a number of our internal systems with agent-first alternatives. We couldn't find anything that worked well as a CRM. After several unsuccessful attempts to model the problem with markdown files and SQLite databases, we realised there was more to it than putting an agent on top of a database. We didn't want to build another dashboard, pipeline view, or data-entry tool. Instead, we wanted the CRM to be something agents can operate directly to create l
RNGD accelerators land in Equinix's Lisbon DCs
Cohere has released Transcribe Arabic, an open-source model for Arabic speech recognition that the company says outperforms Whisper and OmniASR on dialects, code-switching, and bilingual Arabic-English speech. The 2-billion-parameter model is available on Hugging Face under the Apache 2.0 license. The article Cohere Transcribe Arabic is an open-source model built for Arabic's toughest transcription problems appeared first on The Decoder .
While the auto industry wrings its hands over the electric vehicle market, sweating details like aerodynamic efficiency and range anxiety, a new EV startup based in Lisbon, Portugal, is zagging in a different direction. Amble's new electric buggy won't impress anyone with its 0-60 time or its self-driving features (it has none). Instead, it takes […]

<pre style='white-space:pre-wrap;width:81ex'>fix(gateway): bound all-agent usage cache concurrency (#101589) * fix(gateway): bound all-agent usage.cost concurrency to prevent SQLite starvation Replace raw Promise.all in loadAllAgentCostUsageSummary with runTasksWithConcurrency (limit=12), matching the sibling sessions.usage pattern. Unbounded concurrent usage-cost cache reads saturated the SQLite connection pool for deployments with 100+ agents, causing SQLITE_BUSY timeouts or gateway memory cra