Anthropic's 0-Day Discovery. Claude Opus 4.6 found 500+ high-severity vulnerabilities in well-tested open source codebases by reasoning about code the way human researchers would. MORE

The springs of action: What Moves Us, and the Words We Use to Hide It. Bentham's 1817 work reimagined. All actions aim for pleasure or pain avoidance, masked by language. This insight helps decode motives in politics, business, and media. Actions stem from 14 motives like wealth and curiosity. MORE

Terminal Tricks for Agentic Workflows. Hands-on ZSH functions and aliases that enhance Claude Code and other agentic CLI tools. MORE

Jeff Dean on the AI Frontier. From rewriting Google's search stack to Gemini 3 Deep Think distilling to Flash—the history and future of AI at Google. MORE

Cognitive Debt vs Technical Debt. As AI generates more code faster, the real debt accumulates in developers' minds, not in the codebase. MORE

On Discord Alternatives. With Discord requiring age verification, Soatok breaks down why there's no good encrypted alternative today—and what needs to be built. MORE

Peter Steinberger, known recently for his work on OpenClaw, announced he's joining OpenAI to help bring AI agents to mainstream users. Meanwhile, his OpenClaw project is transitioning into an independent foundation to continue its open-source mission. MORE

Bugcrowd Founder Moves On. Casey Ellis shares lessons from 13 years building the bug bounty category and hints at what's next for the hacker mindset. MORE

Obsidian CLI is an early access feature that introduces an command-line interface that lets you control the popular note-taking app from your terminal for automation and scripting. The CLI gives Claude 85% of Obsidian's capabilities with 54x faster searches and 70,000x cheaper token costs compared to traditional file scanning methods. MORE | TEST CASE

Can You Fly That Thing? AI "skills" are programs written in plain English that instantly give AI agents new capabilities, like uploading helicopter pilot knowledge in The Matrix. However, a recent analysis of nearly 5,000 AI repositories found malware hidden in these skill packages, creating serious security risks. MORE

RCE in Google's AI code editor Antigravity. The flaw exploited the browser extension's screen recording feature to write malicious files anywhere on the system, potentially achieving code execution through the Windows startup folder. MORE

XBOW Finds 0-day IDORs. Autonomous security agents discovered two previously unknown IDOR vulnerabilities in Spree Commerce through multi-step reasoning and visual analysis. MORE

Trail of Bits Claude Code Config. Opinionated defaults and security-focused workflows for Claude Code from one of the top security research firms. MORE

Burp AI Agent Extension. A Burp Suite extension adding 53+ MCP tools, AI-assisted analysis, and passive/active scanning across 62 vulnerability classes. MORE

After nearly seven years at HackerOne, Ariel Garcia reflects on building their pentest community from scratch and creating a global hacker ambassador network spanning 45 countries. His biggest achievement was launching the Ambassador World Cup tournament that paid out $2.4 million in its latest edition. MORE

@irsdl - Hacker (ethical), web appsec specialist, trainer, tools builder & apps breaker

@beep - Designer, writer, and tech worker. Started that responsive web design thing.

@rekter0 - Doing things sundialxyz, CTF water_paddler / BlueWater

@HusseiN98D - Entrepreneur, Hacker. WebImmunify, ranked on BugCrowd

@mr_hacker0007 - Bug Hunter intigriti, Ex Bugcrowd, TeamBounters

Font Review Journal. A beautifully designed publication dedicated to in-depth typography reviews and type design analysis. MORE

Gian Galang Portfolio. Striking combat illustration work with bold colors and dynamic compositions worth studying. MORE

Myinstants Soundboard. A collection of trending sound effects and meme sounds. Surprisingly useful for video editing and presentations, but in my case for agentic notifications. MORE

Please start eating fiber. The discussion highlights how many millennials struggle with digestive health issues due to poor dietary habits, sparking widespread conversation about nutrition and wellness in their 30s and 40s. MORE

PhoneLeak: Gemini Data Exfiltration. Security researchers discovered they could exfiltrate data from Gemini mobile by making it call an attacker's phone and encode secrets as DTMF tones. MORE

Moshi: SSH Terminal for Claude Code & AI Agents - Guides and tutorials for AI coding agents and mobile terminal workflows.The complete guide to settin... MORE

Julius: LLM Service Fingerprinting. An open-source tool that identifies what LLM server software is running on any endpoint—Ollama, vLLM, LiteLLM, and 17+ more. MORE

Augustus: LLM Prompt Injection Scanner. Test LLMs against 210+ adversarial attacks covering jailbreaks, prompt injection, encoding exploits, and data extraction. MORE

Showboat and Rodney: Simon Willison's new tools let coding agents build demo documents and automate browser sessions to prove their work actually works. MORE

Modular .bashrc with .bashrc.d/. Source a directory of files instead of one massive .bashrc—a cleaner pattern for shell configuration. MORE

GitButler CLI. A reimagining of git for online-first workflows: parallel branches, stacked PRs without rebase nightmares, and an undo that actually makes sense. MORE

Gwtar: Static Efficient HTML Archives. A clever polyglot format that creates single-file HTML archives that still lazy-load assets via HTTP range requests. MORE

Vulnerability Spoiler Alert. A GitHub Action that alerts on security patches before the CVE drops by monitoring commit messages. MORE

lazyqmd TUI. A terminal UI for qmd files, making command-line documentation navigation easier. MORE

Do Less. Cate Hall discovers that optimization itself can sabotage rest, learning, and meditation—sometimes shutting off the machinery is the real breakthrough. MORE

Nudge Theory for Good Choices. Design your environment so positive behaviors become automatic—practical applications for wealth, health, relationships, and career. MORE

AI Doesn't Reduce Work—It Intensifies It. HBR research suggests AI tools may increase workload rather than decrease it, challenging productivity assumptions. MORE

Stop Optimizing Your Life. Start Living. Dr. K suggests stepping back from endless productivity hacks and self-improvement routines to focus on being present and authentic. MORE

You Could've Invented OpenClaw. Nader Dabit breaks down how to build a persistent AI assistant from scratch in just 400 lines of Python, covering sessions, tools, memory, and multi-channel messaging. He walks through each component step-by-step, showing how you could have invented OpenClaw yourself using basic building blocks like JSONL files, system prompts, and tool execution loops. MORE

The AI Vampire. Steve Yegge warns that AI tools like Claude Code are creating an "energy vampire" effect, draining developers through addictive 10x productivity gains that lead to burnout and overwork. He argues companies are extracting all the value while employees suffer exhaustion, proposing a 3-4 hour workday as the solution to fight back against unsustainable AI-driven expectations. MORE