Microsoft Prompt Shields

Security & Safety

Microsoft Prompt Shields is an Azure AI Content Safety feature that detects jailbreaks and indirect prompt injection attacks hidden in documents or web pages consumed by an agent, generally available since 2024 and extended for agent workloads. It integrates with Azure OpenAI Service and third-party models.

🌐 Visit website ↗

🔗 Similar tools

RAMPART ★ 403

RAMPART is a pytest-native safety and security testing framework for agentic AI, developed…
Vigil ★ 496

Vigil is a security scanner for LLM applications that detects prompt injections, jailbreak…
CrowdStrike Falcon AIDR

CrowdStrike Falcon AIDR, generally available December 2025, provides enterprise-wide visib…
AgentDojo ★ 803

AgentDojo is a research benchmark from ETH Zürich for evaluating prompt-injection attacks …
prompt-firewall ★ 1

prompt-firewall is an early-stage firewall for LLM prompts that detects and blocks prompt …
Inflection 2.5 / Pi

Inflection 2.5 is an empathetic conversational AI model released in March 2024, with its P…

Tools & Infrastructure