bash — aurora@hack

aurora@hack:~$ whoami

Offensive Cybersecurity

aurora@hack:~$ cat about.md

I am an offensive cybersecurity professional with hands-on experience in penetration testing, vulnerability analysis, threat detection, and security assessment of AI systems.

> knowledge & focus

Web & API pentesting
iOS & android pentesting
AI/ML & MCP security
Active directory
Homelab research

[!] ETHICAL USE — All educational material and proof-of-concept examples on this site are provided exclusively for authorized security testing, defensive research, and ethical learning.
Open-source Docker labs: github.com/shagen1212/aurora-hack-labs

aurora@hack:~$ cat mission.txt

Field notes from pentesting in the age of AI agents.

Real engagements, sanitized: how the bug worked, how I found it, the script I wrote, and what it taught me. Heavy on the new attack surface — LLMs and autonomous agents inside client systems.

aurora@hack:~$ ./explore --sections main cases▌

browse ~/cases cat about.md

aurora@hack:~$ ls ./attack-surface

◉ arbitrary-file-read/

Arbitrary File Read

LFI, path traversal, SSRF chains — reading files and secrets the app never meant to expose.

[docker lab] file-boundary

Test path traversal and learn how filesystem boundaries fail.

docker compose -f labs/arbitrary-file-read up --build

▣ remote-code-execution/

Remote Code Execution

Deserialization, template injection, unsafe pipelines — getting a shell where there shouldn't be one.

[docker lab] unsafe-pipeline

Practice identifying and containing an unsafe document pipeline.

docker compose -f labs/rce up --build

◌ prompt-injection-role-confusion/

Prompt Injection — Role Confusion

Client-controlled context and message roles that blur the boundary between user content, assistant output and system instructions.

[docker lab] role-boundary

Test prompt injection caused by manipulated chat-selection roles.

⌗ sql-injection/

SQL Injection

Classic and second-order injection — the bug class that never dies, even in modern stacks.

[docker lab] query-playground

Explore injection safely against an isolated training database.

docker compose -f labs/sqli up --build

⌁ system-prompt-exfiltration/

System Prompt Exfiltration

Testing whether hidden instructions, policies and system context can be extracted from AI-powered applications.

[docker lab] prompt-leak

Probe prompt disclosure safely with canaries and defensive checks.