AI Potluck
Product / UX / Safety & Guardrails

DeepTeam

Confident AI

An open-source red-teaming framework for LLM systems from Confident AI. It simulates adversarial attacks (jailbreaking, prompt injection, multi-turn exploitation) with 50+ built-in vulnerability checks and 20+ attack methods aligned to OWASP Top 10 for LLMs, NIST AI RMF, and MITRE ATLAS, and ships production guardrails.

DeepTeam, Apache-2.0, Confident AI; ~1.9K GitHub stars (June 2026). 50+ vulnerability checks, 20+ attack methods; OWASP/NIST/MITRE aligned. Verified live June 2026.

Openness

5 high confidence
5.0
license
Apache-2.0(OSI)
source
public
self-host
yes
service
none
core-gated
ungated

Fully open source: Apache-2.0 red-teaming framework, self-hostable.

Adoption

2 low confidence
2.0

~1.9K GitHub stars (June 2026); stars-only proxy, capped per methodology.

Capability

4 medium confidence
4.0

Strong standards alignment (OWASP/NIST/MITRE) and broad attack coverage; also ships runtime guardrails.

Unchanged since 2026-08-01 (last edited, not re-checked)