Agentic AI & MCP atlas
16 animated architecture views, MCP message flows, identity, RAG, memory, and cloud security. English · Best on a desktop browser.
Explore the interactive atlas →BASTET / LAB
Security Learning Lab
Visual architecture guides, agentic security experiments, published control mappings, and cybersecurity news.
The framework mapper uses published AWS and Microsoft references (MCSB v2 preview). Cyber news draws from publisher feeds. The breach simulators, self-assessment, and cost calculator remain educational tools; they do not inspect your environment.
Agentic security simulations
12 synthetic experiments. Change controls, replay an attack path, and compare the consequences.
Open all 12 simulations →Agentic AI visual guide
100 illustrated pages: diagrams, worked experiments, Azure/AWS control mappings, and a glossary. PDF · English.
Download the PDF atlas →AWS breach simulator
Educational tool · English
Open tool →Azure breach simulator
Educational tool · English
Open tool →Zero Trust self-assessment
Educational tool · English
Open tool →Framework mapper
AWS & Azure mappings · NIST, CIS, ISO 27001 & PCI DSS · English
Open tool →Latest cyber news
Publisher headlines with source links and publication dates · English
Read the news →Breach cost scenario calculator
Educational tool · English
Open tool →Jump to an agentic security experiment
The agentic simulations use explicit synthetic rules and do not connect to models, credentials, or live systems.
- Break the exfiltration chain
- Catch a cross-tenant RAG leak
- A valid token for the wrong service
- When a “read-only” tool writes
- Containment is not egress control
- Approval and the changed action
- The payment whose reply was lost
- Watch a delegation storm grow
- Spend a finite context budget
- Poison, persist, and delete memory
- A fetch tool crosses the network boundary
- Does the kill switch stop everything?
NEXT STEP
Request a scoping call ↗
Define the scope before you commit.
Tell us about your environment, risk concerns, and target date.