Anonymous, but not unaccountable

Zer0ContextLost is a solo, independent research and writing project covering AI systems — how they fail, how they get attacked, and the tools built to study both. I run the experiments myself, write up what I find, and publish the methodology alongside the results, including the parts that didn't work.

Why anonymous

The writing here is published under a handle rather than a real name by deliberate choice, not to obscure who's accountable for it. Every post is written, tested, and stands behind the same identity every time — zer0contextlost — so the track record accumulates against one name even though it isn't a legal one. The GitHub account linked in the footer carries the code referenced in these posts; nothing here is published without something runnable or reproducible behind it.

What this site is

Mostly security research on AI systems — prompt injection, memory poisoning, model auditing, malware triage tooling — plus occasional posts on tooling and infrastructure that isn't specific to any one finding. Findings that touch real vendors get responsible disclosure before publication where that applies. Nothing here is sponsored, and there's no newsletter or user tracking beyond standard, cookie-free traffic analytics.

Get in touch

For corrections, disclosure follow-ups, or anything else: X / Twitter, GitHub, or zer0contextlost@proton.me.