Overview
What is Mandatez Mcp?
Mandatez Mcp is a neutral trust infrastructure layer for AI agents. It provides cryptographic identity, policy enforcement, human oversight gates, and tamper-proof audit trails across multiple agent frameworks (LangChain, n8n, AutoGen, CrewAI).
How to use Mandatez Mcp?
Install the SDK via npm install @mandatez/sdk, generate an agent identity, create a client with a Supabase URL and anon key, then call client.track() to emit signed events. For framework integration, use the n8n node (n8n-nodes-mandatez) or the LangChain callback (MandateZLangChainCallback).
Key features of Mandatez Mcp
- Ed25519 cryptographic identity per agent
- Policy engine with allow/block/flag rules
- Human oversight gate with Slack/webhook alerts
- Tamper-proof audit trail signed to Supabase
- Compliance report export (HIPAA, EU AI Act, SOC2 coming)
- Works with multiple agent frameworks
Use cases of Mandatez Mcp
- Audit AI agent actions for compliance requirements
- Enforce authorization policies on agent resource access
- Pause agent execution for human approval before sensitive actions
- Generate tamper-proof logs for regulatory or legal proceedings
- Monitor and track agent behavior across heterogeneous frameworks
FAQ from Mandatez Mcp
What does Mandatez Mcp do vs alternatives?
It provides a dedicated trust layer for AI agents with cryptographic identity, policy enforcement, human oversight, and audit trails, integrating across multiple frameworks simultaneously.
What are the runtime requirements?
Requires Node.js, a Supabase project (URL and anon key), and optionally n8n for node-based workflows.
Where does audit trail data live?
All signed events are emitted to a Supabase database configured by the user.
Does Mandatez Mcp support human oversight?
Yes, a human oversight gate can pause execution, send alerts via Slack or webhook, and auto-block on timeout.
Which frameworks are supported?
LangChain, n8n, AutoGen, CrewAI are supported now; Voiceflow, Dify, and Make are planned.
