Vertex Ai MCP Server

🚀 Vertex AI Guide

Vertex AI is a powerful platform for building and deploying machine learning models. This guide will detail how to leverage various tools and services of Vertex AI to optimize your development process.

🚀 Quick Start

📦 Installation

Quick Installation

Use the following command to quickly install the Vertex AI CLI:

bunx vertex-ai-cli --add

Environment Variable Configuration

Create or edit the .env file in the project root directory and add the following content:

AI_PROVIDER=vertex  # Optional values: vertex or gemini
GOOGLE_CLOUD_PROJECT=your_project_id  # Required if AI_PROVIDER is vertex
GEMINI_API_KEY=your_gemini_api_key  # Required if AI_PROVIDER is gemini
VERTEX_MODEL_ID=gemini-2.5-pro-exp-03-25  # Optional, default value from .env.example
GEMINI_MODEL_ID=gemini-2.5-pro-exp-03-25  # Same as above
GOOGLE_CLOUD_LOCATION=us-central1  # Configuration specific to Vertex AI
AI_TEMPERATURE=0.0
AI_USE_STREAMING=true
AI_MAX_OUTPUT_TOKENS=65536  # Default from .env.example
AI_MAX_RETRIES=3
AI_RETRY_DELAY_MS=1000

✨ Features

Toolset

Basic Tools

• vertex-ai-cli: The command-line interface for Vertex AI, used to perform core operations.
• answer_query_websearch: Conduct web searches and answer questions using Vertex AI.
• generate_code: Automatically generate code snippets.

Advanced Features

• Model Fine-tuning: Fine-tune Vertex AI models with custom datasets to improve performance.
• Batch Processing: Support parallel processing of large-scale data to enhance efficiency.
• Real-time Inference: Perform model inference online to obtain immediate results.

Integration and Extension

Integration with Cline

Integrate Vertex AI into Cline (a powerful command-line tool):

1. Open the clib.toml file and add the following content under [mcpServers]:

[vertex-ai-server]
command = "bunx"
args = ["-y", "vertex-ai-cli"]
env = {
    AI_PROVIDER = "vertex",
    GOOGLE_CLOUD_PROJECT = "your_project_id",
    VERTEX_MODEL_ID = "gemini-2.5-pro-exp-03-25",
    AI_TEMPERATURE = "0.0",
    # Other configuration parameters...
}
alwaysAllow = ["answer_query_websearch"]
disabled = false

2. After saving the file, Cline will automatically detect and start the Vertex AI service.

💻 Usage Examples

Development and Debugging

Monitoring Mode

Use the following command to monitor code changes in real-time:

bun run watch

Formatting and Validation

Keep the code style consistent:

bun run format
bun run lint

📄 License

This project is licensed under the MIT License. For details, please refer to the LICENSE file.

By following this guide, you can fully utilize the powerful features of Vertex AI and seamlessly integrate it into your development workflow.