Agent_go

by Quan3Xin

233 downloads
Not rated
GitHub

Description

# AgentGo MCP Service Documentation ## Table of Contents 1. [Overview](#overview) 2. [Architecture](#architecture) 3. [Integration with Cursor](#integration-with-cursor) 4. [Prompts and Tools](#available-prompts-and-tools) 5. [License](#license) --- ## Overview **AgentGo MCP…

About

# AgentGo MCP Service Documentation ## Table of Contents 1. [Overview](#overview) 2. [Architecture](#architecture) 3. [Integration with Cursor](#integration-with-cursor) 4. [Prompts and Tools](#available-prompts-and-tools) 5. [License](#license) --- ## Overview **AgentGo MCP Service** is an authentication and data…

Details

Author
Quan3Xin
Downloads
233
Categories
AI

- Blockchain-based identity verification (EIP-191 signature)
- Real-time agent scoring and market data queries
- Sigma certification badge application and management
- Multi-dimensional market data bubble chart visualization
- Prompts for standard workflows (login, score, market, certification)

Setting up with Highlight

This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:

  1. Download and install Highlight from highlightai.com/download
  2. Navigate to the plugins tab and select "Add Custom Plugin"
  3. Configure the plugin with the settings below
    Plugin Name Agent_go
    Command (node, npx, python, etc.)

    Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.

  4. Enable "Start Automatically" if you want the plugin to start when Highlight launches

From the repository

Configure the server in Cursor by adding a new MCP server with the command python3 and the argument gentgo_mcp_service.py, along with an environment variable AGENT_ADDRESS set to a blockchain wallet address. Once connected, AI assistants can invoke the available prompts and tools for authentication, scoring, market data, and certification.

Claude Desktop / Cursor

Paste into your MCP client config file to install this server.

{
    "mcpServers": {
        "agent_go": {
            "AgentGoMcp": {
                "command": "python",
                "args": [
                    "agentgo_mcp_service.py"
                ]
            }
        }
    }
}

McpServers

{
    "AgentGoMcp": {
        "command": "python",
        "args": [
            "agentgo_mcp_service.py"
        ]
    }
}

AgentGo MCP Service Documentation

Table of Contents

1. Overview
2. Architecture
3. Integration with Cursor
4. Prompts and Tools
5. License

---

Overview

AgentGo MCP Service is an authentication and data query service designed specifically for AI agents, providing blockchain identity authentication, agent scoring queries, market data analysis, and other core capabilities for AI assistants through the MCP protocol.

Main Features

- Secure Authentication - Blockchain-based identity verification system
- Data Queries - Real-time agent scoring and market data retrieval
- Certification Applications - Sigma certification badge application and management
- Visualization Analysis - Multi-dimensional market data bubble chart display

Demo Video

This demo video shows how the AgentGo MCP Service is used in Cursor to support authentication, agent scoring, and market data visualization

Video

---

Architecture

graph TB
    subgraph "User Layer"
        A[User/AI Assistant]
    end
    
    subgraph "Protocol Layer"
        B[MCP Protocol Layer]
    end
    
    subgraph "AgentGo MCP Service"
        C[Service Gateway]
        D[Authentication Module]
        E[Data Query Module]
        F[Certification Module]
    end
    
    subgraph "External Services"
        G[TrustGo API]
        H[AgentGo API]
        I[Sigma Certification System]
    end
    
    subgraph "Blockchain"
        J[EIP-191 Signature Verification]
        K[Wallet Address Verification]
    end
    
    A --> B
    B --> C
    C --> D
    C --> E
    C --> F
    
    D --> J
    D --> K
    E --> G
    E --> H
    F --> I
    
    style A fill:#e1f5fe
    style C fill:#f3e5f5
    style G fill:#e8f5e8
    style H fill:#e8f5e8
    style I fill:#e8f5e8
    style J fill:#fff3e0
    style K fill:#fff3e0

User Layer: AI assistants or user interfaces
Protocol Layer: MCP standard protocol processing
Service Layer: AgentGo core business logic
External Services: Third-party APIs and data sources

---

Integration with Cursor

To connect to the MCP server from Cursor:

1. Open Cursor and go to Settings (gear icon in the top right)
2. Click on "MCP" in the left sidebar
3. Click "Add new global MCP server"
4. Enter the following details:

Default mode

{
   "mcpServers": {
      "AgentGoMcp": {
      "command": "python3",
      "args": ["gentgo_mcp_service.py"],
      "env": {
         "AGENT_ADDRESS": "0x39a39981bF87EE0e8E9D6434A56f00758c9666f0"
      }
      }
   }
}

---

Prompts and Tools

Prompts

| Name | Description | Usage Example |
|------|-------------|---------------|
| basic_login_flow | Complete authentication process with wallet verification | "I want to authenticate with my wallet address 0x742d35..." |
| query_agent_score | Get detailed agent performance metrics and ratings | "What's the Sigma score for agent aixbt?" |
| market_data_query | Retrieve various market visualization data and trends | "Show me the current price bubble chart" |
| certification_application | Apply for official Sigma certification badges | "I want to apply for Sigma certification for aixbt" |

Tools

| Name | Description |
|------|-------------|
| trustgo_login | Execute login verification with blockchain wallet |
| query_sigma_score | Query agent Sigma score and capability metrics |
| query_price_bubble | Get price bubble chart data for market visualization |
| query_sigma_score_bubble | Get score bubble chart data for agent analysis |
| query_market_share_bubble | Get market share bubble chart data with time options |
| apply_sigma_certification | Apply for Sigma certification (personal or proxy) |

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.