Jade Han @windowhan

windowhan.eth | Lead Security Auditor @zellic_io prev. @kalos_security Joined March 2014

Tweets

382
Followers

1K
Following

1K
Likes

1K

V12 @v12sec

2 months ago

🐬🐬🐬 dolphin rce 🐬🐬🐬

41 104 2K 206K 375

View Details

Both my iPad and MacBook that I charged while staying at the hotel ended up having motherboard failures. I’m about to lose a lot of money on repairs. They worked perfectly fine at other hotels, so I have no idea what I’m supposed to do about this. I’m going crazy 🫠

0 0 1 380 0

View Details

Jade Han @windowhan

2 months ago

@d0nutptr @OpenAI Awesome congrats!

0 0 1 199 0

View Details

Kirill @kirillk_web3

5 months ago

instead of watching 2 hours of Netflix tonight, watch this 40-minute masterclass from the founder of a $20B China AI company it's the clearest explanation I've seen of how Agent Swarms and AI systems actually work at scale useful whether you've never built an agent in your life or have been using Claude every day for the past year I took the key ideas and turned them into a practical guide on how to actually build with Kimi find it below

Kirill @kirillk_web3

5 months ago

x.com/i/article/2056…

13 87 569 947K 3K

109 2K 17K 14.0M 55K

View Details

Jade Han @windowhan

5 months ago

@KKDW_KOREA @GONOGO_Korea 우와 이거 결말이 어떻게되엇나요 신박하네요

0 0 0 63 0

View Details

Jade Han @windowhan

5 months ago

@0x15_eth @Zcash @ipwning @ZcashCommGrants Congrats!

0 0 1 175 0

View Details

Jade Han @windowhan

5 months ago

@Krongggggg 동의합니다 추가로 더 대두되는 문제는... 그래서 진짜 취약점이 묻히기더하구요 ㅠ

0 0 1 94 0

View Details

Jade Han @windowhan

6 months ago

@gf_256 I'm always rooting for you. Just ignore the weirdos who are jealous of you. I hope your heart stays unhurt

0 0 1 60 0

View Details

Robert Chen @NotDeGhost

6 months ago

what a ridiculous take. @gf_256 / Zellic have done a ton for the space (speaking as one of their competitors). did you even read the report?

The Lunduke Journal @LundukeJournal

6 months ago

Remember the security firm that Ubuntu hired to audit the (ill-advised, highly buggy) Rust-rewrites of all of the GNU Coreutils? Turns out that security firm is run by @gf_256, who: - Appears to be a man who thinks he's a woman ("trans"). - Uses an anime cartoon of a girl as

386 73 1K 622K 261

0 8 194 19K 7

View Details

Taeyang Lee @5unkn0wn

6 months ago

Everyone focuses on memory corruption bugs in the Linux kernel, but we shouldn’t overlook logical bugs.

Brian Pak @brian_pak

6 months ago

Time to talk about this one. CopyFail (CVE-2026-31431) — a 732-byte Python script that roots every Linux distro shipped since 2017. 🧵

42 449 3K 748K 2K

6 13 124 12K 16

View Details

Jade Han @windowhan

6 months ago

I saw an impressive piece of work someone created for the public good. I'd love to share it with others 😄 kimchicanary.xyz/?lang=en

1 0 3 304 2

View Details

Jade Han @windowhan

6 months ago

@developermin1 관심있습니다.

0 0 0 297 0

View Details

Jade Han @windowhan

6 months ago

@AnticsDecoded oh really? nice work 😄

2 0 1 144 0

View Details

Jade Han @windowhan

6 months ago

@devkoriel 익명으로 하심이...

1 0 0 328 0

View Details

Jade Han @windowhan

6 months ago

@devkoriel 공개적인 sns에 이렇게말씀하셔도되나요...ㄷㄷ 걱정되는..

1 0 0 5K 0

View Details

Jade Han @windowhan

7 months ago

@AnticsDecoded @immunefi No! Only manual audit 🙂

0 0 1 95 0

View Details

Jade Han @windowhan

7 months ago

I just found a confirmed bug on @immunefi #immunefitribe immunefi.com/s/ss/?severity… I achieved it after a very long debate/argument And... yet another thing that got buried 🙃

6 1 89 3K 6

View Details

Jade Han @windowhan

7 months ago

When doing bug hunting with Claude Code, I tried using various good skills and tools to improve my workflow, but the results were not as great as I expected. After thinking about it, part of the issue was that I wasn't doing great prompt engineering or providing the right context — but also, since it's basically a black box in terms of what actually gets sent to and received from the LLM Provider, I had no idea what the problems were or which direction to improve things. So I built a separate tool for debugging purposes, and I figured there might be others out there with the same need, so I'm sharing the repo: github.com/windowhan/clau… (vibe-coding 100%) Features I'm thinking of adding later: - Intercepting requests being sent to the LLM Provider so a human can intervene mid-process and adjust the direction - Routing specific requests or specific sub-agent actions to a different LLM Provider - Attaching a management Agent to monitor and supervise whether Claude Code has sufficient context while running, and check if anything has been missed if anyone have additional ideas, feel free to me <3