Start Here / The Map

How deep does the rabbit hole go?

This is the map. Krypteia Sec tracks agentic offensive AI security: autonomous hackbots, MCP and agent audits, daily research. Five levels, shallow to deep. Pick where you start.

See the map Or jump to today's intel

The Choice

This is your last chance to look away

The blue pill

Close the tab. The agentic-AI threat keeps moving without you.

The red pill

Stay. We show you how to break AI systems, and how to defend them.

Show me how deep

The Map

Five levels, shallow to deep

Everything on this site, grouped by how far down you want to go. Start at the top for the daily read, or drop straight to the deep end.

Level 1

Daily Signal

For anyone who just wants today's read on AI threats.

Threat Intel

63 daily threat briefs. Source-cited, TLP:CLEAR.

AI Briefs

59 daily briefs tracking the agentic-AI frontier.

Argos Feed

A live feed of AI security content as it lands.

Level 2

Get Oriented

For newcomers who need the words and the basics first.

Glossary

The AI security vocabulary. Look up any term you hit.

Guides

Plain field guides on AI attack surfaces, FAQ-backed.

The Operator

Who runs this site, and why it stays anonymous.

Level 3

Learn the Craft

For people ready to study, foundations through red team.

Learn

30 structured lessons, foundations through red team tracks.

Academy

A curated AI education hub across multiple tracks.

Level 4

Run the Plays

For practitioners who want to do it with their hands.

AI Hacking Playbook

16 live plays mapped to OWASP LLM Top 10 and MITRE ATLAS.

Field Manual

Chain Claude Code features as a red-team operator. Five phases.

Level 5

The Deep End

For readers who want the full body of research.

Research

120 long-form articles on AI agents, hackbots, and LLM attacks.

Now You Know The Way

Want the work behind the map?

See what we do, how we work in our approach, or reach us directly.

Or email us at info@krypteiasec.com