AI Security Matrix

A curated list of AI-enabled security testing tools.

Definitions

Category

agent 26 Runs its own task loop, deciding the next step from what it just found.
mcp 14 Serves tools to a model over Model Context Protocol. The model drives.
scanner 23 Runs a fixed set of checks and reports what they found.
skill 19 Prompts and instructions for a model. No program of its own.

Scope

ad 13 Active Directory: on-premises domain identity.
agentic 12 AI agents and the loops they run, as the thing under test.
api 8 HTTP and RPC interfaces behind an application.
binary 10 Compiled executables and firmware.
cloud 2 Provider accounts and their configuration.
code 13 Source, in a repository or a diff.
container 0 Images, registries, and what runs inside them.
entra 3 Entra ID: Microsoft's cloud identity.
llm 12 The model itself: its prompts, its guardrails, and what it emits.
reporting 3 The record of a test and what it becomes: coverage, findings, the evidence behind them, and the report.
mobile 3 iOS and Android applications.
modfile 5 Serialised model files, whose formats can carry executable code.
network 23 Hosts, ports, and traffic.
recon 23 Finding what exists, before anything is tested.
redteam 12 Adversary emulation across a whole estate.
social 1 People: phishing, pretexting, and what they hand over.
webapp 34 Browser-facing applications.