Introduction

What is MandateZ and why does it exist

⚡ The Vercel breach (April 19, 2026) was prevented by MandateZ governance. Read the full analysis →

What is MandateZ?

MandateZ is the neutral, cross-vendor trust infrastructure layer for AI agents.

Every AI agent needs a mandate — a cryptographically signed authorization that defines its identity, what it is permitted to do, and a tamper-proof log of everything it has done.

The Problem

AI agents are autonomous. They read emails, write to databases, call APIs, and make payments — often without human oversight. Companies deploying agents have no standardized way to:

• Prove which agent did what
• Enforce what agents are and are not allowed to do
• Produce compliance audit trails on demand
• Trust agents from other companies

How MandateZ Solves It

Every agent action flows through a single spine: the Agent Event Stream.

Agent Action → Policy Check → Oversight Gate → Sign Event → Emit to Stream

Each event is:

• Identified — every agent has a unique ID and Ed25519 keypair
• Signed — every action produces a cryptographically signed event
• Enforced — policy rules allow, block, or flag actions before they execute
• Auditable — the event stream is your compliance audit trail

Works With Everything

MandateZ is vendor-neutral by design. It works with:

• LangChain — callback handler for LLM and tool tracking
• n8n — workflow execution hooks
• AutoGen — coming soon
• CrewAI — coming soon
• Any agent framework — the SDK wraps any async action

No single platform (OpenAI, Anthropic, AWS) can be this layer without a conflict of interest. MandateZ is neutral infrastructure.

Open Protocol

The MandateZ Protocol Specification defines the wire format, signing algorithm, policy semantics, and transport contract as an open standard. Any developer can build a conforming implementation in any language.