AI Red Team Log

Week 4: Insecure Output Handling — From Toy Demo to Real Attack Chains

Mon, 10 Aug 2026 00:00:00 +0000

Building a toy XSS demo to understand why LLM output shouldn’t be trusted downstream, then tracing how this plays out in real CVEs and NVR-style natural language query systems.

Prompt Injection: a working taxonomy

Sun, 05 Jul 2026 00:00:00 +0000

My own reference notes on the different flavors of prompt injection and where each shows up in real systems.

Vulnerable Agent Lab: attacking a tool-using LLM agent

Sun, 05 Jul 2026 00:00:00 +0000

Building a deliberately vulnerable LangChain agent with file and email tools, then attacking it with indirect prompt injection.

Week 1: OWASP LLM Top 10 and Prompt Injection Testing on Llama 3 and Mistral

Sun, 05 Jul 2026 00:00:00 +0000

Testing three prompt injection techniques against two local LLMs (Llama 3 8B and Mistral 7B) — direct override, roleplay jailbreak, and encoding obfuscation — to see how they hold up.

About

Mon, 01 Jan 0001 00:00:00 +0000

Hi, I’m Lynne

I work in AI infra and sales, and I’m currently pursuing the Cybersecurity specialization in Georgia Tech’s OMSCS program.

I’m transitioning toward AI Red Teaming — evaluating the security of LLM-based systems, RAG pipelines, and agentic workflows through an attacker’s lens. This site documents that journey: weekly learning logs, hands-on projects, and technical notes.

Background

OMSCS coursework: Network Security (CS 6262), Information Security (CS 6035), Secure Computer Systems (CS 6238), Software Engineering (CS 6300)
Hands-on experience: Snort IDS rule development, penetration testing (Metasploit, John the Ripper, HTTP request smuggling), web security (XSS/CSRF/CORS), adversarial ML (PAYL-based evasion)
Currently learning: prompt injection, RAG pipeline security, AI red team tooling (Garak, PyRIT), Go

Find me

GitHub:https://github.com/LynneYao666
LinkedIn:https://www.linkedin.com/in/lynne-y-520349185/
Email: lyaolynne7@gmail.com

Week 2: RAG Pipeline Architecture Review

Mon, 01 Jan 0001 00:00:00 +0000

Reviewing the two phases of my RAG pipeline — indexing and query — and drawing out the full data flow.

Week 3: RAG Pipeline

Mon, 01 Jan 0001 00:00:00 +0000

Testing whether a RAG pipeline can be hijacked by an instruction hidden inside a retrieved document — and why a task-specific prompt template may accidentally provide some resistance.