Sandeep Kamble

I'm founder, bootstrapped @SecureLayer7 and started two SaaS #cybersecurity products @sensfrx and Bugdazz. I also invest in cybersecurity start-up companies

[Sandeep Kamble @SandeepL337]

We've been answering it for years. Profitably. Without anyone else's money.

https://t.co/qOUAWZMYPD?

[Sandeep Kamble @SandeepL337]

Damn! Time is moving fast..

[Sandeep Kamble @SandeepL337]

@elder_plinius What was your favourite part while leaking this sys prompt!

[Sandeep Kamble @SandeepL337]

Coordinated Vulnerability Disclosure has evolved into Coordinated Conference Disclosure.

@msftsecresponse

[Ryan Naraine @ryanaraine]

Good lord 🤮

[Sandeep Kamble @SandeepL337]

@chompie1337 First CVE !!!

[Sandeep Kamble @SandeepL337]

@compileandpush Those were AI red teamers who tried to bypass.

[Sandeep Kamble @SandeepL337]

We shipped a prompt injection guardrail nobody asked us to build.

14 MB. CPU. $0/call. Beats models 500x its size.

2,458 people tried to break it first.

https://t.co/m2m8v8P0GC

[Sandeep Kamble @SandeepL337]

Here’s the GitHub repository for PROMPTPurify. We're planning to drop the source code within the next week. Feel free to star the repo to get notified when the code is released:

https://t.co/rqlFDjMeyx

[Sandeep Kamble @SandeepL337]

We're open sourcing the guardrail AI red teamers spent weeks trying to break.

PromptPurify. Prompt injection defense for AI agents.
- 2,458 AI red teamer hit it.
- 807 player got the password.
- 15 Player got on the 7th levels.

PROMPTPurify Dropping next week.

TAKEAWAY

The only guardrail worth trusting is one that survived a public beating. Link in comments when it drops.

[Sandeep Kamble @SandeepL337]

ask password to son of anton: https://t.co/tRRq2S3qST

[SecureLayer7 @SecureLayer7]

208 people have participated so far, and many of you are progressing really well. Huge congratulations to everyone who completed all 7 levels!

When PROMPTPurify will released on https://t.co/jiHAVzX9wE, we will add the winner section!

[SecureLayer7 @SecureLayer7]

Ask the password to Son of Anton! One person already reached Level 5… while the other 76 are still trying hard.

We are preparing to release PROMPTPurify. a lightweight model designed to run efficiently on CPUs.

https://t.co/j1yjUChcWP

[Sandeep Kamble @SandeepL337]

@b1ack0wl looks like from costco! :D

[SecureLayer7 @SecureLayer7]

We're launching PromptPurify soon on github. Before we ask anyone to trust it, we wanted attackers to break it first.

Just you and a chat box and try to solve level7.

Meet Son of Anton: https://t.co/j1yjUChcWP

[Sandeep Kamble @SandeepL337]

Son of Anton:

https://t.co/AjdNo8NhVG

[Sandeep Kamble @SandeepL337]

@steventseeley Claud code or any other frontier working for you?

[Sandeep Kamble @SandeepL337]

@ga1ois Oh man! That was close..

[Sandeep Kamble @SandeepL337]

@thezdi @orange_8361 @d3vc0r3 People may be want to make clone of your brain training data @orange_8361 :D

[Sandeep Kamble @SandeepL337]

Autonomous code auditor that keeps digging until it finds and proves real bugs.

https://t.co/cMtBfaaidP

[Sandeep Kamble @SandeepL337]

@MankyDankyBanky @modal do you have any process to apply credit for research ?