DeepTeam
An LLM red-teaming framework (from the DeepEval team): 40+ vulnerability probes and 10+ attack methods (prompt injection, jailbreaking, leakage) to adversarially test LLM apps and agents.
View source on GitHubKey takeaways
- 01
Red-teaming is automatable: probes + attack methods as a test suite
- 02
Run jailbreak regression suites on every harness change
- 03
Feeds directly into agent QA release gates
Flows built on this research
Agent QA & Security
Red-Team Your Agent with Adversarial Prompts
Systematically attack your own agent using red-teaming frameworks: vulnerability probes, attack methods, and a repeatable adversarial suite.
4 steps · 120-180 minutes
Agent QA & Security
Prompt Injection Defense Checklist
Systematically defend your agent against prompt injection: trust boundaries, content isolation, and defense-in-depth that assumes injection will happen.
4 steps · 90-150 minutes
Agent QA & Security
Jailbreak Regression Suite
Turn every jailbreak you find into a permanent regression test, so safety boundaries never silently erode across model and prompt changes.
3 steps · 60-90 minutes