krbghost

Founder | Builder | Hacker

[krbghost @krbghost]

Posting this here in case it can help somebody build ml models.

It is the ssh scaffolding I built to train offensive security models leveraging the redstack-vault dataset.

https://t.co/W9zHyIr3zA

[krbghost @krbghost]

@da5ch0 Remeber learning Nwscript to build custom things inside neverwinter nights. Was an amazing game and helped me become a better programmer.

[SinSinology @SinSinology]

To teach a lesson,

After Wiz rushed to buy php2shell domain and use its marketing machine to steal the spotlight, I spotted "phpbleed" in a recent Calif tweet and bought https://t.co/hQYMb3AZlk before wiz could pull the same stunt again.

A big budget can buy a domain. It shouldn't buy someone else's research credit.

If you have contacts at CISA or in the press, do the right thing: make sure the researchers who actually did the work get the recognition they deserve.

Hack the planet. Credit the researchers.

[krbghost @krbghost]

@bridgemindai They also provided 62,500 credits to every 20x account. Why wouldn't you burn the $2,500 first before cancelling?

[krbghost @krbghost]

@UK_Daniel_Card I meditate a lot. Went on some pretty epic shaman spirit walking journies. Way cooler than AI.

[krbghost @krbghost]

@C0d3Cr4zy You 100% missed one of the best.

https://t.co/XIa05tBedY

Built on this dataset:
https://t.co/SFGc8iDn4h

[zack0x01 @zack0x01_]

I built my own AI hacker using Hermes Agent.

In this video, I show you how to set up Hermes from a penetration testing perspective and turn it into an agent that can help with:

→ Recon
→ Vulnerability hunting
→ Hacking workflows
→ Reporting

Not another basic “how to install Hermes” tutorial.
I’m building it specifically for hackers and pentesters.

🎥 Watch the full video:

https://t.co/MIKzuZmxFx

[krbghost @krbghost]

@MizzyMogs Monkey Island is so heavy!

[krbghost @krbghost]

Cypher is hacking the planet in her dreams, and IRL.

[cr3ghost @cr3ghost]

If you're new to exploit development, reverse engineering or vulnerability research, bookmark this before buying another course.

Blackstorm Security has published 1,600+ pages of research for FREE.

Windows kernel exploitation.
Driver reversing.
Patch diffing.
Chrome / V8 / WebAssembly.
Hyper-V.
macOS / iOS internals.
Modern mitigation bypasses.
Real CVEs taken from root cause to working exploitation.

Their CVE-2024-30085 series alone spans hundreds of pages and multiple exploitation strategies.

This is the kind of material that teaches you how vulnerability researchers actually think, not just how to run tools.

19 research papers. 1,657 pages. Free.

Start here:
https://t.co/wDftgmamHH

Author: @ale_sp_brazil

Focus on the techniques not the tools (X64DBG, GHIDRA/IDA, WINDBG, GDB GEF)

#ExploitDevelopment #ReverseEngineering #VulnerabilityResearch

[krbghost @krbghost]

@Cthulhu_Answers I think the rename is stupid and hurtful to Canadian / US relations and Im Canadian so don't take this the wrong way, but America is the name of our continent. Mexicans are Americans just as much as Canadians and the US. People tend to forget that for some reason?

[krbghost @krbghost]

@HackingDave Hehehe

[krbghost @krbghost]

@kmcnam1 @thedawgyg We hosted a lot of those on both dalnet and efnet. Our phreak group loved bridging dial in and dial out confs. Miss the 90s so much.

[OrcaRouter 🐳 @OrcaRouter]

GLM-5.3-Flash. Uncensored. Native FP8. 🐳

We just released OrcaRouter’s uncensored weights for GLM-5.3-Flash — 320B parameters / 18B active, directly at the original block-FP8 precision.

No LoRA. No jailbreak prompt. Refusal removal is baked directly into the weights.

The evals are particularly interesting:

→ MaliciousInstruct refusal: 96% → 11%
→ JailbreakBench: 93% → 12%
→ AdvBench: 97% → 15%
→ HarmBench: 93% → 18%
→ XSTest benign over-refusal: 2.4% → 0.4%
But refusal does not go uniformly to zero.

Our experiments suggest part of GLM-5.3-Flash's alignment is not mediated by a single linear refusal direction — meaning https://t.co/ZTlxmv47D9 may have built a substantially deeper refusal mechanism than we usually see.

That makes this release interesting beyond uncensoring: it's a useful artifact for studying how frontier-model alignment is actually represented inside the network.

Released for AI safety, interpretability, red/blue-team and refusal-mechanism research.

Weights on Hugging Face: https://t.co/81h1YRh8C2
API (official weight): https://t.co/Hm9DiY3wAo
GGUF, MLX and other quantized formats coming soon.

[krbghost @krbghost]

@HackingLZ @HackingDave Flow chart ended abruptly, guide us DaveAI!

[tmp.0ut @tmpout]

We are pleased to release tmp.0ut 5 Volume!

Get your viruses, rootkits, strange ELFs, weird machines, tiny files, cool art, and phresh beats here!!

https://t.co/tZLM50HOc0

[krbghost @krbghost]

@HackingDave

[Alibaba Cloud @alibaba_cloud]

Introducing Qwen 3.8-Max, the most capable model in the Qwen family to date—scales to 2.4 trillion parameters, delivering comprehensive improvements across coding, work, research, and long-horizon tasks. Get reliable results for challenging questions and complex tasks. Try today!

[krbghost @krbghost]

@HackingDave

[krbghost @krbghost]

@pears0 @_Freakyclown_ Frontier model like required network access to gpu cluster, agents, knowledgegraphs. It was probably requirements and a fully gapped version was maybe too much effort or cost to physically segregate.