Skip to content
View nuryslyrt's full-sized avatar
🌟
Discover Stars!
🌟
Discover Stars!

Highlights

  • Pro

Block or report nuryslyrt

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
nuryslyrt/README.md

> whoami

root@nur-gucu:~$ whoami --verbose

A decade-deep offensive security practitioner who stumbled into the AI frontier and never looked back. I've spent years breaking things; networks, applications, assumptions, and now I do it to the models that are reshaping the world.

I think loudly, build quickly, and hack relentlessly. My blog, ThinkingTokens, is where those three instincts collide.


Tip

Rules of engagement: I break models to make them stronger. If something I publish helped you patch, ship, or think — that's the whole point. DM me the war story.


> ps aux | grep nur

currently running in background:

PID USER STAT %CPU COMMAND
0001 nur R< 98% orpheus --skill-based-orchestration --daemon
0042 nur R 84% adversarial_fuzzer --target=frontier-llms --loop
1337 nur R 71% thinkingtokens.ai --publish --chaotic-mode
2026 nur S 55% red_team_ops --client=redacted --engagement=active
3141 nur S 33% python -m tool_builder --never-finished
9999 nur Z 2% sleep_schedule (defunct)

R = running • S = sleeping • R< = high-priority • Z = zombie, don't ask


> cat ~/.arsenal

tools



🔴 offensive

LLM jailbreaksprompt injection model extractionRAG poisoning MCP abuseagentic escape multi-agent topology attacks

🟣 research

adversarial prompting failure cartography skill engineering contracts emergent threat modeling

🟢 build

Python + automation fuzzers that nobody asked for ORPHEUS orchestration SKILL.md everything


> systemctl status thinking.service

most recent transmissions from thinkingtokens.ai:

● thinking.service - Loaded (chaotic-good; enabled; caffeine-triggered)
   Active: active (writing) since forever ago

📡 Introducing ORPHEUS: Skill-Based AI Orchestration #orchestration #skills #claude-code #agentic-systems

📡 The SKILL.md Wild West: Deep-Dive Across 9 Coding Agents #agents #llm #ai-tooling

📡 Your Skill Is a Behavioral Contract. Are You Sure It Holds? #skill-engineering #vibe-coding #agentic-systems

📡 Every AI Model Breaks — Red Teaming Is Failure Cartography #red-teaming #llm #adversarial

📡 The Ouroboros Effect: Multi-Agent Exploits in the Wild #multi-agent #adversarial #topology

📡 The Art of Adversarial Prompting #red-teaming #prompting


> nmap -sV -T4 nur.gucu

Starting Nmap ( you're already running it, you're on this page )
Host is up (0.001s latency).

PORT     STATE      SERVICE        VERSION
22/ssh   open       ssh            LinkedIn (banner: "Nur Yesilyurt")
80/http  open       blog           ThinkingTokens v2026
443/tls  open       dm             X / Twitter (@nuryslyrt)
1337/*   open       collaboration  accepting adversarial-ai work
5150/*   filtered   inbox          recruiter spam dropped at firewall
9001/?   open       camping        status: frequently afk in the wild

Service Info: OS: chaotic_thinker_linux; Uptime: 10+ years; Honeypots: yes

sudo ./unlock_easter_egg.sh  (click me — every red teamer leaves a backdoor)
[*] escalating privileges............ ok
[*] brute-forcing readme............. ok
[*] locating hidden flag............. ok
  ░█▀▀░▀█▀░█▀▀░░░▀█▀░█░█░█▀▀░█▀█░█░█░▀█▀░█▀█░█▀▀
  ░█░░░░█░░█▀▀░░░░█░░█▀█░█▀▀░█░█░█░█░░█░░█░█░█░█
  ░▀▀▀░░▀░░▀░░░░░░▀░░▀░▀░▀▀▀░▀░▀░▀▀▀░▀▀▀░▀░▀░▀▀▀

         flag{you_read_the_whole_readme_hire_me}

     thanks for scrolling. if you made it this far,
     you're exactly the kind of person I like working with.
     ping me. bring coffee. or a broken model.
                                        — nur

> wall -a "let's connect"

channel protocol address latency
🧠 blog HTTPS thinkingtokens.ai real-time
💼 professional mTLS linkedin/nuryesilyurt < 24h
🐦 shitposts & research XMPP/X @nuryslyrt chaotic
🏕️ irl line-of-sight somewhere off-grid with a tent signal lost

CVE-2026-NUR  •  "exposure to this profile may cause an uncontrolled urge to jailbreak a model you were supposed to be defending"

Severity: informational  •  CVSS: 10.0 (aesthetic)  •  Patch: N/A — working as intended



Pinned Loading

  1. AttackImaginator AttackImaginator Public

    LLM Supported Attack Scenario Creator from Code Review

    Python 14

  2. AISecTips-Tricks AISecTips-Tricks Public

    AI Red Team & Blue Team Tips & Tricks!

    10 1

  3. ORPHEUS ORPHEUS Public

    Python

  4. SKILLBridge SKILLBridge Public

    Cross Compatible Skill Creation, Auditing and Converting for Coding Agents