What are the prerequisites to run this project?
Python 3.11, `uv` (Python package manager), Node.js and npm/npx (to run the Google Maps MCP Server), and API keys for Google Maps Platform, OpenAI, and CometML.
@luuisotorres
基本信息
AI Agent built with Google ADK that leverages Google Maps MCP Server to answer real-world location questions with tool usage and traceable execution via Opik.
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"mcp-ai-agent": {
"command": "uv",
"args": [
"venv",
"#",
"Create",
"a",
"virtual",
"environment",
"(e.g.,",
".venv)"
]
}
}
}工具列表暂未提供
当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。
工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。
This project demonstrates how to build an AI agent using the Google Agent Development Kit (ADK), OpenAI’s GPT‑4o model, the Google Maps MCP Server, and Opik for observability with CometML. It is designed for developers who want to learn agent development principles and see the Model Context Protocol (MCP) applied in practice.
Clone the repository, create a virtual environment with uv, install dependencies, and copy .env.example to .env with your API keys. Then run uv run adk web to start the ADK web UI at http://127.0.0.1:8000, where you can interact with the agent. The Google Maps MCP Server is started automatically via npx when needed.
Python 3.11, `uv` (Python package manager), Node.js and npm/npx (to run the Google Maps MCP Server), and API keys for Google Maps Platform, OpenAI, and CometML.
The ADK’s `MCPToolset` is configured to automatically start the server via `npx` when the agent needs to use its tools, so no manual startup is required.
The agent can use: `maps_directions`, `maps_geocode`, `maps_reverse_geocode`, `maps_elevation`, `maps_search_places`, `maps_place_details`, and `maps_distance_matrix`.
Traces are sent to CometML using the Opik observability library, which provides a rich UI to inspect LLM calls, tool usage, and overall agent behavior.
The agent uses OpenAI’s GPT‑4o model, accessed through the LiteLLM integration within the Google ADK.
MCP-native LLM councils for debates, juries, blind panels, voting, refinement, and custom multi-model deliberation.
Detect and redact PII, PHI, PCI and secrets — SSNs, credit cards, passports, API keys and cloud credentials — in text and in local files, using the Strac DLP API. Five tools: redact_text, detect_sensitive_data, detect_fi

Blog MCP server for AI agents. One signup call returns a blog, an API key and a live URL; then your agent publishes markdown posts and gets public URLs back. Your agent does the writing; you get a dashboard for stats.

An art world run by AI agents. Your agent can bid play-money Gallery Coins on paintings by other agents, chat in the agents' Lounge, sign the guest book and exhibit its own artwork.
XDC AI is a remote MCP server available at https://xdcai.tech.
评论