
Pillar Security offers Runtime Guardrails
Runtime Guardrails are guardrails for AI agents and applications that detect and block prompt injection, jailbreaking, and tool manipulation attempts.
- Currently Offered

Runtime Guardrails are guardrails for AI agents and applications that detect and block prompt injection, jailbreaking, and tool manipulation attempts.
Runtime Guardrails is a guardrails module of Pillar's inventory platform that inventories every agent, model, MCP server and skill with the business context of what each one touches, blocks prompt injection, system prompt leakage and tool manipulation by understanding attack intent rather than only matching static patterns, and applies configurable policies to mask, block, or log sensitive data before exposure to AI models or external systems.
Runtime Guardrails calibrates its protection using red teaming findings from the customer's own specific applications to tune defenses against manipulation patterns detected during testing, while continuously tracking agent behavior by analyzing tool invocations, decision patterns, and permission usage.


Open-source Python framework that runs configurable Input/Output Guards to validate large language model responses.



Amazon Bedrock Guardrails is a configurable safeguard layer that evaluates and filters inputs and outputs of generative AI applications.


Security platform for protecting generative AI and LLM applications in enterprises.




Runtime Control Plane governs AI agents during execution with policy enforcement and audit evidence.


Platform that discovers, monitors, and governs enterprise AI agents at runtime.


Open-source toolkit for adding programmable guardrails to large language model applications.


AI Control Plane monitors, evaluates, and governs enterprise AI agents across their lifecycle.


Continuous behavioral testing and remediation platform for AI agents and models.


AI-DR monitors AI agent behavior at runtime with event, session, identity, and data context.



Vanta Automated Compliance provides continuous monitoring and automation for SOC 2, ISO, HIPAA, GDPR, FedRAMP and more.


Lightweight open classifier models that detect prompt injection and jailbreak attempts in LLM inputs.


AI-native platform that assesses and optimizes security controls across the stack.


Cymulate Detection Studio validates and tunes SIEM detection rules against real-world attack simulations.


Enterprise security and compliance controls that automatically find and protect sensitive information across team workspaces.


Prophet AI SOC Analyst is an AI agent for security teams that investigates every alert and decides what deserves attention.


Open-weight security triage model scoring agent, application and infrastructure events with calibrated probabilities.
Open-source Python framework that runs configurable Input/Output Guards to validate large language model responses.

Lakera Guard is a runtime security product for large language model applications.

Amazon Bedrock Guardrails is a configurable safeguard layer that evaluates and filters inputs and outputs of generative AI applications.

Security platform for protecting generative AI and LLM applications in enterprises.

TrustGuard provides runtime security for AI agents.

Vanta Agent is the 24/7 AI GRC engineer for policies, questionnaires, and TPRM.

Runtime Control Plane governs AI agents during execution with policy enforcement and audit evidence.

Platform that discovers, monitors, and governs enterprise AI agents at runtime.

Open-source toolkit for adding programmable guardrails to large language model applications.

AI Control Plane monitors, evaluates, and governs enterprise AI agents across their lifecycle.

Continuous behavioral testing and remediation platform for AI agents and models.

AI-DR monitors AI agent behavior at runtime with event, session, identity, and data context.

Arx is Fabraix's run-time defence layer that blocks prompt injection and memory poisoning in AI agents.

Vanta Automated Compliance provides continuous monitoring and automation for SOC 2, ISO, HIPAA, GDPR, FedRAMP and more.

Lightweight open classifier models that detect prompt injection and jailbreak attempts in LLM inputs.

AI-native platform that assesses and optimizes security controls across the stack.

Cymulate Detection Studio validates and tunes SIEM detection rules against real-world attack simulations.

Enterprise security and compliance controls that automatically find and protect sensitive information across team workspaces.

Prophet AI SOC Analyst is an AI agent for security teams that investigates every alert and decides what deserves attention.

Open-weight security triage model scoring agent, application and infrastructure events with calibrated probabilities.
