Mandatez Mcp

Name: Mandatez Mcp
Author: mandatez

Local

core-dashboard-black.vercel.appSecurity1Updated 22w ago

Repository Visit website

About Mandatez Mcp

MandateZ provides identity, policy enforcement, human oversight, and auditing infrastructure for teams that need to govern and trace AI agent actions. It signs events with Ed25519 and includes n8n and LangChain integrations.

Key features

01Generate Ed25519 cryptographic identities for agents
02Allow, block, or flag actions with policies
03Pause execution for human oversight
04Sign and record agent action events
05Export compliance reports (in development)
06Integrate with n8n and LangChain

Use cases

01Set rules to allow or block agent actions
02Block execution after an oversight timeout
03Create signed audit events for workflow runs
04Review audit events for agent activity

#security #ai-agents #audit-trail #trust #compliance #owasp #ed25519 #cryptography

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config

{
  "mcpServers": {
    "core-mandatez": {
      "command": "npx",
      "args": [
        "-y",
        "@mandatez/mcp"
      ],
      "env": {
        "SUPABASE_URL": "<SUPABASE_URL>",
        "SUPABASE_ANON_KEY": "<SUPABASE_ANON_KEY>",
        "MANDATEZ_OWNER_ID": "<MANDATEZ_OWNER_ID>"
      }
    }
  }
}

Run commandnpx -y @mandatez/mcp

Environment variables to set

SUPABASE_URL

SUPABASE_ANON_KEYSecret

MANDATEZ_OWNER_ID

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages

npm@mandatez/mcpstdio

Tools

Saved tool list · updated 3w ago

Parameter	Type	Required	Description
name	string	Yes	Human-readable agent name
metadata	object	No	Optional metadata to attach to the agent

Details

From the project README

MandateZ

Every agent needs a mandate.

MandateZ is the neutral trust infrastructure layer for AI agents. Cryptographic identity, policy enforcement, human oversight gates, and tamper-proof audit trails — for every agent, across every framework.

Why MandateZ

AI agents are autonomous. They read files, call APIs, send emails, make payments. Nobody can prove what they did, whether they were authorized, or produce a compliance trail on demand.

MandateZ fixes all four. It works with LangChain, n8n, AutoGen, CrewAI, and every other framework — simultaneously.

Install

npm install @mandatez/sdk

Quickstart

import { generateAgentIdentity, MandateZClient } from '@mandatez/sdk';

const identity = await generateAgentIdentity();
const client = new MandateZClient({
  agentId:         identity.agent_id,
  ownerId:         'your_org_id',
  privateKey:      identity.private_key,
  supabaseUrl:     process.env.SUPABASE_URL!,
  supabaseAnonKey: process.env.SUPABASE_ANON_KEY!,
});

const event = await client.track({
  action_type: 'read',
  resource: 'emails',
});
// → signed, validated, emitted to your event stream

What you get

Ed25519 cryptographic identity per agent — unique ag_ prefixed IDs with keypair signing
Policy engine — allow/block/flag rules with wildcard resource matching
Human oversight gate — pause execution, alert via Slack/webhook, auto-block on timeout
Tamper-proof audit trail — every action signed and emitted to Supabase
Compliance report export — HIPAA, EU AI Act, SOC2 (coming)

Integrations

n8n

npm install n8n-nodes-mandatez

Drop the MandateZ node into any workflow. Every execution gets a signed audit event.

LangChain

import { MandateZLangChainCallback } from '@mandatez/sdk';

const callback = new MandateZLangChainCallback(client);
const chain = new ChatOpenAI({ callbacks: [callback] });

More coming

AutoGen, CrewAI, Voiceflow, Dify, Make.

Architecture

Everything flows from one spine: the Agent Event Stream.

Agent Action → Policy Engine → Oversight Gate → Sign (Ed25519) → Emit to Stream

Every surface — SDK, dashboard, compliance engine, directory — reads the same stream. One data layer. No duplication.

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

Read the getting-started guide

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.