FeedCity

This series of articles chronicles the history, both real and pseudo, behind Gabriel Knight 3: Blood of the Sacred, Blood of the Damned. It is possible to trace the Plantard family tree a fair ways back without relying on the Lobineau dossier, but not as far back as the time when the Merovingian kings ruled […]

Pictured above is the new DeskPi Super4C installed in an 8U mini rack. The Super4C is a 4-node Raspberry Pi CM5 cluster board that solves two pain points I had with the older Super6C.

I was testing this board around the same time I helped kick off the SBCC 2026, the Single Board Cluster Competition for students. A dozen or so university teams squared off to run the best mini HPC cluster with a budget of $6,000, and a couple days to benchmark six HPC workloads.

What to do when upstream ghosts you

Codex CLI 0.128.0 adds /goal

The latest version of OpenAI's Codex CLI coding agent adds their own version of the Ralph loop: you can now set a /goal and Codex will keep on looping until it evaluates that the goal has been completed... or the configured token budget has been exhausted.

It looks like the feature is mainly implemented though the goals/continuation.md and goals/budget_limit.md prompts, which are automatically injected at the end of a turn.

Via @fcoury

Tags: ai, openai, prompt-engineering, generative-ai, llms, coding-agents, system-prompts, codex-cli, agentic-engineering

I wrote a simple script that translates text at the command line, completely offline. Here’s an example of how it works on my computer: echo '¿Cómo estás?' | translate # => How are you? It combines a few tools: TranslateGemma, a special-purpose language model for trans...

Our evaluation of OpenAI's GPT-5.5 cyber capabilities

The UK's AI Security Institute previously evaluated Claude Mythos: now they've evaluated GPT-5.5 for finding security vulnerability and found it to be comparable to Mythos, but unlike Mythos it's generally available right now.

Tags: ai, openai, generative-ai, llms, anthropic, claude, ai-security-research, gpt

It's a common misconception that we can't tell who is using LLM and who is not. I'm sure we didn't catch 100% of LLM-assisted PRs over the past few months, but the kind of mistakes humans make are fundamentally different than LLM hallucinations, making them easy to spot. Furthermore, people who come from the world of agentic coding have a certain digital smell that is not obvious to them but is obvious to those who abstain. It's like when a smoker walks into the room, everybody who doesn't smoke instantly knows it.

I'm not telling you not to smoke, but I am telling you not to smoke in my house.

— Andrew Kelley, Creator of Zig

Tags: zig, llms, ai, generative-ai

We need RSS for sharing abundant vibe-coded apps

Matt Webb:

I would love an RSS web feed for all those various tools and apps pages, each item with an “Install” button. (But install to where?)

The lesson here is that when vibe-coding accelerates app development, apps become more personal, more situated, and more frequent. Shipping a tool or a micro-app is less like launching a website and more like posting on a blog.

This inspired me to have Claude add an Atom feed (and icon) to my /elsewhere/tools/ page, which itself is populated by content from my tools.simonwillison.net site.

Tags: atom, matt-webb, rss, ai, vibe-coding

acetaminophen vs. ibuprofen

Today's links How not to ban surveillance pricing: Maryland's new consumer protection law is all loophole. Hey look at this: Delights to delectate. Object permanence: Google's 8,000 Linux servers; Battleshoe; Knitted potholes; Unpack the court; "Robot Artists and Black Swans"...

A Brazilian tech firm that specializes in protecting networks from distributed denial-of-service (DDoS) attacks has been enabling a botnet responsible for an extended campaign of massive DDoS attacks against other network operators in Brazil, KrebsOnSecurity has learned. The firm's chief executive says the malicious activity resulted from a security breach and was likely the work of a competitor trying to tarnish his company's public image.

Though wind and solar continue to carve out larger and larger shares of world energy supply, the modern world still runs on petroleum, and will continue to do so for the foreseeable future.

If you followed along with the recent joyful celebrations of the Artemis cruise around the moon, and took a moment to dive into the photographic archives of the mission, you might have noticed that all of the original images were shared by NASA on the venerable photo sharing...

Zig has one of the most stringent anti-LLM policies of any major open source project: No LLMs for issues. No LLMs for pull requests. No LLMs for comments on the bug tracker, including translation. English is encouraged, but not required. You are welcome to post in your nati...

Release: llm 0.32a1

• Fixed a bug in 0.32a0 where tool-calling conversations were not correctly reinflated from SQLite. #1426

Tags: llm

Release: llm 0.32a0

See the annotated release notes.

Tags: llm

After a busy March, April was a little quieter. But don’t worry, I still have a bunch of little links for you to click on. Things I published GitHub’s uptime hasn’t been great recently. Even though I dislike the Microsoft subsidiary, I wrote “In defense of GitHub’s poor up...

I just released LLM 0.32a0, an alpha release of my LLM Python library and CLI tool for accessing LLMs, with some consequential changes that I've been working towards for quite a while. Previous versions of LLM modeled the world in terms of prompts and responses. Send the mod...

Support for remote controlling Windows PCs may be added to Raspberry Pi Connect, Raspberry Pi's free remote access service.

When they announced Pi Connect in 2024, I speculated the service was launched in response to RealVNC's sluggish adoption of Wayland, leading to Pi users lacking a solid remote access solution after Pi OS 12 'Bookworm' was launched.

The service was helpful for those who had one or more Raspberry Pis to access, but the Pi Connect daemon didn't run on Windows or macOS at the time, so a true competitor to RealVNC (at least for basic use cases) it was not.

It's shocking how much you can deduce about what the labs are doing from a handful of equations and a blackboard