Prompt Injection Protection

MCP Defender is an open-source AGPL-3.0 secure proxy that scans and filters MCP tool calls in real time, using LLM detection and deterministic signatures to block prompt injection, tool poisoning, credential theft, arbitrary code and remote commands.

Lakera protects generative‑AI and LLM deployments with real‑time threat detection, sub‑50 ms latency, and safeguards against prompt injection, data leakage, and jailbreaks. It offers workforce monitoring, granular policy controls, red‑team vulnerability simulation, and multilingual multimodal suppor

Prompt Engineering Institute offers resources for mastering prompt engineering and AI applications, including free courses, tutorials, and community support. Its 5C framework aids in crafting effective prompts and optimizing AI model performance through advanced techniques.

Prompt Attack offers a diverse marketplace of customizable prompts, including calligraphies, icons, and illustrations for various projects like SEO research and web development. Access trending prompts and exclusive midjourney content for optimal results.

iDox.ai Guardrail is an AI security tool that intercepts data on-device to prevent sensitive information from reaching external AI models like ChatGPT. It blocks risky prompts and enforces centralized policies with audit trails for compliance across finance, healthcare, and government sectors.

Protex AI is an AI-powered workplace safety platform that uses computer vision on existing CCTV to detect safety risks in real-time. It provides predictive analytics, automated reporting, and compliance workflows to help EHS teams prevent incidents and improve operational safety.

Sinaptic.AI is a browser extension that detects and blocks sensitive personal data (PII/PHI) in real-time using a hybrid engine. It prevents accidental data exfiltration at the network level and provides visual feedback, custom rules, and enterprise management for compliance.