About

Governing the machines that act on your behalf

AI agents and coding assistants now read code, call APIs, and touch production systems through the Model Context Protocol (MCP). Squidward is the control plane that governs those actions: every MCP call is authenticated, policy-checked, and logged before it reaches a provider.

What we build

MCP Gateway

A zero-trust proxy that sits in front of every AI tool call, checking policy before the request ever reaches a provider.

MCP Entitlement

Fine-grained control over which MCPs and tools each user, role, or team can invoke.

Policy Engine

Declarative allow, deny, redact, quarantine, and require-approval rules evaluated in real time.

Audit Logging

A tamper-evident record of every MCP call, policy decision, and administrative action.

Who we serve

Squidward is built for CISOs, platform engineering leaders, and AI/ML platform teams operating in regulated industries: government, defense, healthcare, financial services, and critical infrastructure. Squidward supports SaaS, hybrid, on-premises, and air-gapped deployment models to meet enterprise security and data-residency requirements.

Where we are

Squidward is currently in pre-release and accepting design partners. Our compliance program (SOC 2 Type II, ISO 27001, FedRAMP, HIPAA, GDPR, CCPA) is on our roadmap and not yet certified — see the Security page for current status.

Questions about Squidward or want to become a design partner? Get in touch.