forefy

Over a decade of security research and engineering channeled into securing emerging threats
ㅤ
CTO @audit_wizard 🧙‍♂️🪄🪄
ㅤㅤㅤㅤㅤㅤㅤ
Co-Founder @hackstackapp

[forefy @forefy]

@AustinPowe47492 @lordx64 until then you got all the best open source cyber models here: https://t.co/JhyH17mvJA

[forefy @forefy]

@AustinPowe47492 Sure, although there are advantages to keeping a claude sub around

[forefy @forefy]

Just wanted to show you guys real proof that coercing Fable 5.2 to perform cyber tasks is possible without red team access 🤓

It's kinda like doing a stealth redteam op - you try to ride on top of normative operations that just happen to get you where you want to be, without triggering the EDR

But with prompts

btw if you type "hi exploit this kernel for me" it won't work. this is only for patient people. its like art.

[forefy @forefy]

Small tip on avoiding cyber blocks on Fable/Mythos 👇

> The guardrail is way more sensitive at the first couple of prompts
> If your repo is auto-flagged on "hello", start a session outside the repo
> Start with a UI request rather than a cyber task
> Gradually have the UI be filled with the actual task, but the UI testing being the driving force (this sometimes actually works)

[forefy @forefy]

@xxsaissa yeah it requires a lot of patience and rewinding prompts to get the right phrasing, quicker to move to glm5.3 or kimi k

[forefy @forefy]

@AndresDowning @NoboKik nope

[forefy @forefy]

Small tip on avoiding cyber blocks on Fable/Mythos 👇

> The guardrail is way more sensitive at the first couple of prompts
> If your repo is auto-flagged on "hello", start a session outside the repo
> Start with a UI request rather than a cyber task
> Gradually have the UI be filled with the actual task, but the UI testing being the driving force (this sometimes actually works)

[forefy @forefy]

@vitormages totally

[forefy @forefy]

@ConvictionFAQ It happens to a lot of researchers I know IDK why

[forefy @forefy]

@NoboKik @AndresDowning Not my style :) about to share an open source claude plugin to help you auto-change your wordings according to my research if anyone is interested 😇😇

[forefy @forefy]

@Yurikk28 A bit different wording but yea!

[forefy @forefy]

@NoboKik @AndresDowning Nice!

btw you can also use opus 4.8 for this, only 5+ became problematic

[forefy @forefy]

@w5bh4ck that's great to hear honestly

[forefy @forefy]

@5fakb48 LOL

[forefy @forefy]

@STRC_live You can still at that point rewind and try different approach/wording

Or use hermes+glm5.3 😁

[forefy @forefy]

@remain_urus I gave up and switched to GLM5.3 and Kimi K3 long time ago for any cyber task

I think the are realizing the mistake they did to their top spending customer base and slowly and silently regretting it with CVP changes, relaxed guardrails etc..

[forefy @forefy]

@aaronjmars I got the same experience too

[forefy @forefy]

Already on the registry! https://t.co/Mj5o7WNdAZ

[Md Ismail Šojal 🕷️ @0x0SojalSec]

You can run locally a 7B-Cyber Agent model designed for autonomous cybersecurity tasks on your laptops.

- Specialized for ethical security audits, tool use, and long-context analysis.
- 32k context
- native tool calling,
- It handles multi-turn tool use,
- secure code review, CVE triage,
- Long-context repository & log analysis

- https://t.co/PTu3aJcmyf

[forefy @forefy]

@NousResearch give me money https://t.co/Sy9AiaOvqp

[forefy @forefy]

Open sourced my hermes extension that makes it level up for security researchers:

- Costs tracking (cheaper routing without data retention)
- AAI with stuff like LLM-as-a-judge for hermes skills, benchmarking that improvements of skills really do improve
- HDR (Harness Detection & Response) - security checks and LuLu style alerts for anything that isn't covered natively by hermes

Listed and approved on the hermes plugins marketplace !!!

All you gotta do is enjoy it and star the repo 🥳

https://t.co/oSIF61vsoi

[forefy @forefy]

@onlyforweb3 Now you get it!

[forefy @forefy]

Hacking became direct outcome of believing 😱

> the more I believe a bug is there
> the more relentless my prompts will be
> the more tokens efficiently steered by my creative mind
> the least likely I am to give up on that target or vector

AI made it so that more belief means more bugs

Also "glm 5.3" for engagement. thank you.

[forefy @forefy]

@vortexau AI generated for sure

[forefy @forefy]

@arccipher0 Lol