Donato Capitella

Introducing spikee, an open-source toolkit for testing LLM applications against prompt injection attacks that can lead to exploitation—like data exfiltration, XSS, and resource exhaustion. Easily create custom datasets to match your specific use cases. spikee.ai

@Ysrthgrathe42 @TeksEdge @Kyuz0 @TeksEdge can you please update the tag to my handle on here, @dcapitella ?

@TeksEdge @Kyuz0 @dcapitella Got a good link to the work? His account shows suspended currently.

🚀 New MTP Support Dropped for Strix Halo! Qwen3.6-35B-UD-Q4_K_XL = ~76 tps. @kyuz0 updated his AMD Strix Halo Toolboxes with full MTP support. ✅ ~2x faster token generation ✅ Works with Qwen 3.6 models ✅ New ROCm 7.2.3 toolbox ✅ Pre-built containers (llama.cpp + vLLM) No more manual setup pain. Just pull & run.

@Ysrthgrathe42 @TeksEdge @Kyuz0 well, that's not my account... on here I am simply @dcapitella. You can find all the information on strix-halo-toolboxes.com.

@Flo408800717920 @TeksEdge @Kyuz0 Llama.cpp, deliberately with 1 concurrent request.

@TeksEdge @Kyuz0 @dcapitella Good news. Will test later. Bench is with llama or vllm ? CAN it handle 2 parallel requests ?

@know2whyi1ft @TeksEdge @Kyuz0 Not by a lot and not on all model architecture and quants. This is mostly a llama.cpp implementation difference, they don't have optimized kernels for ROCm for all architectures and quants.

@TeksEdge @Kyuz0 @dcapitella Vulkan is faster than ROCm is very embarrassing for AMD.

@_mohansolo @antigravity Finally some communication!

Thanks for all of the Antigravity feedback over the last couple of days, especially around the IDE. Our intention was never to remove the IDE support for developers, and we should have been clearer with that in the product from the beginning. We’ve made it clearer in 2.0 on how to connect to the IDE, fixed issues with opening the IDE on Windows machines, provided instructions to restore IDE settings & extensions, and more. New releases for the Antigravity IDE and Antigravity 2.0 have rolled out with these changes. We should have done better so we’re going to reset everyone’s Gemini quota for the week again.

@giovanniJurado You using my latest toolboxes? Have you got llama-cockpit?

My Strix Halo spiritual guide. Although I still can't get any MTP speed upgrade... I must be doing something wrong.

@robzolkos I have just released a new video with pi performance on local ahrdware (AMD Strix Halo / R9700) with Qwen 3.6: youtu.be/JyS8A-5LIY8?si…, also including full SWE Verified Mini runs, for people to get an idea of speed of completion and accuracy of the quants we can run locally.

Hey Pi-curious peeps - or people of Pi - there are now 79 Pi videos, tutorials, full courses etc on the video page lazypi.org/videos.html Good way of learning all about it and switch from coding harnesses tied to model providers that do not care about you....

In this video, I look at what works when running coding agents locally on Strix Halo and Radeon R9700. I cover pi and opencode running Qwen 3.6, and SWE-bench mini benchmark results: youtu.be/JyS8A-5LIY8

@CACandChill Charge 10x what we do today to fix bugs :)

Software engineers, imagine AI becomes 10x better at coding in 3 years. what are you doing next?

@MrOnyszko @antigravity It was cheap and good at the beginning, and I like the Windsurf feeling. Then the price went up and quotas became impossible to work with, plus they just don't have capacity even for paying users. I'll likely be moving away unless the new 3.5 flash model is good

@dcapitella @antigravity Curious. Why do you use it?

@antigravity - already fully broken the second time I launch the new Antigravity 2.0. Login flow is totally broken.

@Bob_Welton I think so, but I'm not a Windows user, so I'd recommend searching for tutorials online.

@dcapitella Great video. But is there a working method of doing this with Windows/Vulkan in the Strix Halo ?

youtu.be/MI0Pm1d6YF4 MTP can accelerate LLM inference 2x, especially for coding agents. This video covers what MTP is and the performance improvements you can expect for Qwen 3.6 on AMD Strix Halo & Dual Radeon 9700.

@baldwinblueprnt lol - nice joke :)

Run it as an administrator

@antigravity How is Antigravity 2.0 missing a search feature for projects and conversation!?!?!?!?!

@menticapti @Hyperoptic I think they should rebrand

@dcapitella @Hyperoptic Paying £45.00 a month for dial-up at this stage. My god.

@Hyperoptic having another outage within 1 month ...

@antigravity is extremely poor at session management, Many times I re-open it and my last session is totally erased. This is basic stuff, we're not getting anywhere close to AGI if the thing can't even persist a session on disk...

My two cents on how to interpret the Mozilla / Mythos findings. linkedin.com/pulse/donato-c…

Salerno airport is so disorganised, not able to handle border control for the ONLY flight landing. Making people wait 30 minutes to go through border control when there is only one flight