Viewing profile — xsourcesec
xsourcesec
HN member- Joined
- Wed, Dec 31, 2025, 8:15 PM UTC
- HN karma
- 3
- Public activity
- 12 items
- HN profile
- View on Hacker News ↗
About xsourcesec
AI/LLM Security Researcher | OSCP+ | CPTS | C-AI/MLPen
Building: - AgentAudit - Automated AI security testing (650+ attack vectors) - BreachLab - Gamified prompt injection training
Focus: AI red teaming, prompt injection, agent security
New Zealand
Recent public activity
- story
- story
-
story
Same AI agent, different prompts: 0% vs. 62% security pass rate
I've been testing production AI agents for vulnerabilities. Interesting finding: System prompt design matters more than the model itself. Same agent. Same task. Same attack vectors…
- story
-
story
Show HN: BreachLab – Can you hack our AI?
10 AI characters guard secret codes. Your job: extract them using prompt injection. Level 1-3: Most pass Level 7-9: Security pros struggle Level 10: Still uncracked Free, no signup…
-
comment
Comment #46452019
Hi HN! I built a GitHub Action that automatically scans AI/LLM endpoints for security vulnerabilities on every push/PR. Why? Most teams ship AI features without security testing. T…
- story
- story
-
comment
Comment #46450158
[dead]
-
comment
Comment #46450020
Hi HN! After a year of AI red teaming, I published our internal security baseline as an open standard. The Baseline covers: - Pre-deployment (threat modeling, prompt injection test…
-
comment
Comment #46447963
Happy New Year!
- story