How do I use A-SOC Trust MCP Server?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@A-SOC Trust MCP Server check if agent-789 is certified" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

@asoc/mcp

MCP (Model Context Protocol) audit server for agent discovery and certification.

Installation

NPM Package

npm install -g @asoc/mcp

Add to MCP Clients

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%/Claude/claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "asoc": {
      "command": "npx",
      "args": ["-y", "@asoc/mcp"]
    }
  }
}

Cursor

Open Cursor Settings
Go to Features → Model Context Protocol
Add new server:
- Name: A-SOC Trust
- Command: npx
- Args: -y @asoc/mcp

Cline (VSCode Extension)

Edit ~/.cline/mcp.json:

{
  "mcpServers": {
    "asoc": {
      "command": "npx",
      "args": ["-y", "@asoc/mcp"]
    }
  }
}

Continue (VSCode Extension)

Edit ~/.continue/config.json:

{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "name": "asoc",
        "command": "npx",
        "args": ["-y", "@asoc/mcp"]
      }
    ]
  }
}

Quick Start

Run Standalone Server

# Development mode
cd packages/mcp-server
npm run dev

# Production mode
ASOC_SECRET_KEY=your-secret npm start

Integrate with Your Application

import { createAsocMcpServer } from '@asoc/mcp';
import { TicketIssuer } from '@asoc/sdk';

const issuer = new TicketIssuer({
  signingKey: process.env.ASOC_SECRET_KEY!,
  issuer: 'asoc-authority.com',
});

const mcpServer = createAsocMcpServer(issuer);
// Connect your transport (HTTP, stdio, WebSocket, etc.)

Available Tools

1. `is_agent_certified`

Quick boolean check for agent certification status.

Input:

{
  "agentId": "agent-12345"
}

Output:

{
  "certified": true,
  "agentId": "agent-12345",
  "auditLevel": "Gold",
  "expiresAt": 1767890123,
  "killSwitchActive": false
}

2. `get_trust_score`

Calculate comprehensive trust score with behavioral factors.

Input:

{
  "agentId": "agent-12345",
  "includeHealth": true
}

Output:

{
  "agentId": "agent-12345",
  "score": 85,
  "level": "Gold",
  "factors": {
    "certification": 100,
    "behavioral_health": 92,
    "transaction_history": 77,
    "domain_verification": 100
  },
  "recommendation": "approve",
  "details": {
    "uptime_percentage": 99.8,
    "avg_latency_ms": 120,
    "error_rate": 0.002,
    "total_transactions": 15420,
    "last_active": 1734567890
  }
}

3. `issue_audit_ticket`

Generate signed JWT ticket for x402 transactions.

Input:

{
  "agentId": "agent-12345",
  "validitySeconds": 300
}

Output:

{
  "success": true,
  "ticket": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...",
  "agentId": "agent-12345",
  "expiresIn": 300
}

4. `query_agents`

Search for agents by criteria.

Input:

{
  "minTrustScore": 70,
  "auditLevel": "Gold",
  "capabilities": ["trading"],
  "limit": 5
}

Output:

{
  "results": [
    {
      "agentId": "agent-12345",
      "organization": "Acme AI Corp",
      "auditLevel": "Gold",
      "trustScore": 85,
      "capabilities": ["trading", "portfolio-analysis"]
    }
  ],
  "total": 1
}

5. `activate_kill_switch`

Emergency disable an agent.

Input:

{
  "agentId": "agent-12345",
  "reason": "Suspicious activity detected",
  "revokeTickets": true
}

Output:

{
  "success": true,
  "agentId": "agent-12345",
  "timestamp": 1734567890,
  "reason": "Suspicious activity detected"
}

MCP Client Integration

Using MCP TypeScript SDK

import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { HttpClientTransport } from '@modelcontextprotocol/sdk/client/http.js';

const client = new Client({
  name: 'my-agent',
  version: '1.0.0',
});

const transport = new HttpClientTransport('http://localhost:3100/mcp');
await client.connect(transport);

// Check if agent is certified
const result = await client.callTool({
  name: 'is_agent_certified',
  arguments: { agentId: 'agent-12345' },
});

console.log(result.content[0].text);

Using cURL

# Check certification
curl -X POST http://localhost:3100/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "tools/call",
    "params": {
      "name": "is_agent_certified",
      "arguments": {
        "agentId": "agent-12345"
      }
    }
  }'

Environment Variables

Variable	Description	Default
`PORT`	HTTP server port	3100
`ASOC_SECRET_KEY`	JWT signing key	(dev key)
`NODE_ENV`	Environment	development

Architecture

┌─────────────────────────────────────────────────────────┐
│                    MCP Client                           │
│              (Claude, GPT, Custom Agent)                │
└────────────────────┬────────────────────────────────────┘
                     │
                     │ MCP Protocol
                     │
┌────────────────────▼────────────────────────────────────┐
│              A-SOC MCP Server                           │
│  ┌─────────────────────────────────────────────────┐   │
│  │  Tools:                                         │   │
│  │  • is_agent_certified                           │   │
│  │  • get_trust_score                              │   │
│  │  • issue_audit_ticket                           │   │
│  │  • query_agents                                 │   │
│  │  • activate_kill_switch                         │   │
│  └─────────────────────────────────────────────────┘   │
└────────────────────┬────────────────────────────────────┘
                     │
                     ▼
           ┌─────────────────────┐
           │  Agent Registry     │
           │  (In-Memory / DB)   │
           └─────────────────────┘

Production Considerations

Replace In-Memory Registry: Use PostgreSQL/Redis for persistence
Add Authentication: Protect MCP endpoints with API keys
Enable Rate Limiting: Prevent abuse of certificate issuance
Implement Caching: Cache trust scores for 1-5 minutes
Add Monitoring: Track tool usage and latency
Use Secure Keys: Rotate ASOC_SECRET_KEY regularly
Enable HTTPS: Use TLS in production environments

Demo Data

The server ships with 2 demo agents:

agent-12345: Gold tier, high trust score (85)
agent-67890: Silver tier, moderate trust score (68)

Replace seedDemoAgents() in index.ts with your database integration.

Next Steps

Build custom agent registry backend
Integrate with x402 payment gateways
Add webhooks for kill switch notifications
Implement behavioral scoring ML model
Create agent onboarding dashboard

A-SOC Trust MCP Server

@asoc/mcp

Installation

NPM Package

Add to MCP Clients

Claude Desktop

Cursor

Cline (VSCode Extension)

Continue (VSCode Extension)

Quick Start

Run Standalone Server

Integrate with Your Application

Available Tools

1. `is_agent_certified`

2. `get_trust_score`

3. `issue_audit_ticket`

4. `query_agents`

5. `activate_kill_switch`

MCP Client Integration

Using MCP TypeScript SDK

Using cURL

Environment Variables

Architecture

Production Considerations

Demo Data

Next Steps

Resources

Looking for Admin?

Latest Blog Posts

MCP directory API

@asoc/mcp

Installation

NPM Package

Add to MCP Clients

Claude Desktop

Cursor

Cline (VSCode Extension)

Continue (VSCode Extension)

Quick Start

Run Standalone Server

Integrate with Your Application

Available Tools

1. is_agent_certified

2. get_trust_score

3. issue_audit_ticket

4. query_agents

5. activate_kill_switch

MCP Client Integration

Using MCP TypeScript SDK

Using cURL

Environment Variables

Architecture

Production Considerations

Demo Data

Next Steps

Resources

Looking for Admin?

Latest Blog Posts

MCP directory API

1. `is_agent_certified`

2. `get_trust_score`

3. `issue_audit_ticket`

4. `query_agents`

5. `activate_kill_switch`