AI Agent Security Audit | CAIN-42

One system, three sitescainstudio.online mcpgate.online clawx.click

CAIN-42 CAIN Studio

Platform

Unified Trio Hub Inventions Invention Atlas Architecture Runtime & Trust Fabric Governance Trust State Blast Radius & Impact Trust Drift Live Cluster System Status Metrics Console

Enforce

MCPGate Exfiltration Killswitch MCP Explorer MCP Security Scanner Guard (decorator)Sandbox & Egress Tool Catalog Skills & Plugins Tools Integrations Policies

Developers

Quickstart Documentation API Reference SDK Docs AI Quickstart MCP Docs SDK Python SDK TypeScript SDK CLI Learn Troubleshooting FAQ Glossary

Proof

TLA⁺ Specification Models Live Now Proof Center Verify Yourself Conformance Formal Verification Independent Testing Byzantine Cluster Changelog Roadmap What CAIN Does Not Do Threat Model Security

Observability

Observability Traces Trajectory Governed Memory Context Governance Evaluation Budget & Metering Sentinel (Moat 7)Resilience Agent Security

Enterprise

Compliance EU AI Act Cyber Insurance Marketplace Plans Pricing DPA SLA Subprocessors Services

Company

Mission Founder The Company Live Demo Book a Demo Join / Careers llms.txt

Sign in Get started

Menu

How it works Technology Proof FAQ Docs Pricing Sign in

This offer has moved. The AI Agent Security Audit starts at $5,000:
see the audit offer and the live product demos.

Our mission

Make autonomous software accountable.

Our goal is that every consequential action an autonomous system takes passes through a boundary that is able to refuse it, and leaves behind a record that survives the incident review months later. Not a dashboard that describes what already happened.

Read the full mission and what we hold ourselves to →

Platform

Architecture
Trust domains
How a decision is made
Supported runtimes
Service catalog

Deployments

CAIN Studio — hosted
MCPGate — self-hosted
Pricing
MCPGate (self-hosted)

Developers

Quickstart
API reference
Glossary
Changelog
Status

Trust & security

Trust center
Security & disclosure
Threat model
Sub-processors
What we don't do yet

Company

Mission
Founder
Roadmap
Contact
Report a vulnerability

Legal

Terms
Privacy
Refunds
Data processing (DPA)
Service levels

CAIN Trust Fabric — CAIN Studio — hosted deployment. Operated by FIRME AI CORPORATION, 2021 Fillmore St · San Francisco, CA 94115 · United States of America . © 2026 FIRME AI CORPORATION. All rights reserved. Phone: 1-844-CAIN420.

Status security.txt API reference