Day 14. The CVSS 10.0 advisory I filed against the @omidotme backend on April 15 is still in Triage. Zero engagement. The maintainer closed the public PR that fixed nine of the fourteen issues two days ago without merging. Three new findings shipped in the same window.
This is a coordinated disclosure writeup published on Day 14 after a private report. It is not a call to exploit. The post is here, the GHSA link inside is still private until BasedHardware moves it: github.com/kasparovabi/secur…
Open-sourced a Claude Code skill set that brings agentic discipline
(Plan-Execute, Reflexion, Self-Refine, Ralph Loop) into reusable
protocols. Inspired by Claude Mythos Preview behaviors.
Honest scope: ~40-60% of the gap, no magic.
🔗 github.com/kasparovabi/claud…
I just created a simple AI Video Generator App using LumaLabs new Ray-2 model API.
It is made with the help of Gemini 2.0 Pro Experimental. And it works smoothly, you can directly download the output video into your computer easily.
I test the OpenAI’s new o3-mini model through a unique project. A real-time ant colony simulation that showcases emergent optimization in action.
In the simulation, ants forage for food, deposit dynamic pheromone trails that evolve over time, and naturally optimize their paths between the nest and food sources. The visual elements, powered by OpenGL and Pygame on my macOS M1 Pro, bring these behaviors to life in a captivating 2D environment.
It is fascinating how advanced AI models can drive innovative simulations and deliver insights into complex, real-world behaviors. I’m excited about the potential of these technologies and the new perspectives they offer.
Video to 70's Cartoon AI exploration. Created with #AnimateDiff and #IPAdapter to stylize input video for a retro cartoon look.
🎥 Result (10fps) vs input (30fps iPhone footage)
#aivideo#stablediffusion#ComfyUI
More imagery (and result video by itself) in thread!🧵