SecurityCodexClaudeChatGPT

AI Agent Red Team

Stress-test autonomous agent behavior and tool use.

Difficulty

Advanced

Risk

High

Helps with
Test shell misuse
Probe secret handling
Attack browsing flow
Use cases
Agent safety
Tool permissions
Pre-launch AI QA
Inside
SKILL.mdIncluded
ScriptsNot listed
ReferencesIncluded
AssetsNot listed
agentred teamtool use
Checks
HighSKILL.md

Check before using

Confirm credentials and deployment targets.
Run dry checks before publishing.
Review generated scripts, permissions, and rollback path.
Source

Curated pack

Source

Similar skills

🧨Prompt Injection Lab

Attack your own AI features before users do.

🕵️Threat Modeling

Map assets, trust boundaries, and abuse paths.

🔐Secrets Check

Inspect code and config for accidental credentials exposure.

🪪Auth Flow Review

Review login, sessions, redirects, and permission edges.

🛡️Security Best Practices

Framework-aware security review for JS, Python, and Go.

🧮Permission Matrix

Map roles, actions, and access edge cases.