Skip to main content
Glama
deenesjakoruzh
OpenGerwin

mcp-google-agent-platform-docs

by OpenGerwin
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

mcp-google-agent-platform-docs

为 AI 智能体提供 Google AI 平台文档的 MCP 服务器。

Python 3.10+ MCP License: MIT

OpenGerwin MCP Servers 的一部分

这是什么？

一个 MCP (Model Context Protocol) 服务器，让 AI 智能体能够直接访问 Google 的 AI 平台文档 — 包括当前的 Gemini Enterprise Agent Platform (GEAP) 和旧版的 Vertex AI Generative AI 文档。

您的 AI 助手无需再对 API 细节进行幻觉猜测，而是可以实时查阅实际的文档。

Related MCP server: mise-en-space

功能特性

  • 🔍 全文搜索：涵盖 3400 多页文档

  • 📄 按需获取：根据需要下载并缓存页面

  • 🗂️ 双重来源：当前的 GEAP + 旧版 Vertex AI 文档

  • ⚡ 智能缓存：72 小时 TTL，网络错误时使用过期缓存作为后备

  • 🗺️ 自动发现：通过站点地图扫描（每周）发现新页面

  • 🧩 即插即用：适用于 Claude Desktop、Cursor、VS Code 以及任何 MCP 客户端

快速入门

安装

# Using pip
pip install mcp-google-agent-platform-docs

# Using uv (recommended)
uv pip install mcp-google-agent-platform-docs

配置 Claude Desktop

添加到您的 claude_desktop_config.json：

{
  "mcpServers": {
    "google-agent-platform-docs": {
      "command": "mcp-google-agent-platform-docs"
    }
  }
}

配置 Antigravity (Google)

添加到 ~/.gemini/antigravity/mcp_config.json：

{
  "mcpServers": {
    "google-agent-platform-docs": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-google-agent-platform-docs",
        "run",
        "mcp-google-agent-platform-docs"
      ]
    }
  }
}

配置 Cursor / VS Code

添加到您的 MCP 设置中：

{
  "mcpServers": {
    "google-agent-platform-docs": {
      "command": "mcp-google-agent-platform-docs",
      "transport": "stdio"
    }
  }
}

工具

search_docs

通过关键词搜索文档。

search_docs("Memory Bank setup", source="geap")
search_docs("function calling", source="vertex-ai")

get_doc

获取特定页面的完整内容。

get_doc("scale/memory-bank/setup", source="geap")
get_doc("multimodal/function-calling", source="vertex-ai")

list_sections

浏览文档结构。

list_sections(source="geap")

list_models

所有可用 AI 模型（Gemini、Imagen、Veo、Claude 等）的快速参考。

list_models()

文档来源

来源 ID

平台

页面数

状态

geap

Gemini Enterprise Agent Platform

2300+

主要 (当前)

vertex-ai

Vertex AI Generative AI

1100+

旧版 (归档)

GEAP 章节

  • Agent Studio — 可视化智能体构建器

  • Agents → Build — 运行时、ADK、Agent Garden、RAG 引擎

  • Agents → Scale — 会话、记忆库、代码执行

  • Agents → Govern — 策略、智能体网关、模型防护 (Model Armor)

  • Agents → Optimize — 可观测性、评估、质量警报

  • Models — Gemini、Imagen、Veo、Lyria、合作伙伴、开源模型

  • Notebooks — Jupyter 教程

配置

用于自定义的环境变量：

变量

默认值

描述

MCP_DOCS_CACHE_DIR

~/.cache/mcp-google-agent-platform-docs

缓存目录

MCP_DOCS_CONTENT_TTL

72

页面缓存 TTL (小时)

MCP_DOCS_STRUCTURE_TTL

7

结构缓存 TTL (天)

MCP_DOCS_DEFAULT_SOURCE

geap

默认文档来源

MCP_DOCS_HTTP_TIMEOUT

30

HTTP 超时时间 (秒)

开发

# Clone
git clone https://github.com/OpenGerwin/mcp-google-agent-platform-docs.git
cd mcp-google-agent-platform-docs

# Install dependencies
uv sync

# Run server locally
uv run mcp-google-agent-platform-docs

# Test with MCP Inspector
uv run mcp dev src/mcp_google_agent_platform_docs/server.py

架构

mcp-google-agent-platform-docs/
├── sources/                    # YAML source configurations
│   ├── geap.yaml               # GEAP (primary)
│   └── vertex-ai.yaml          # Vertex AI (legacy)
├── src/mcp_google_agent_platform_docs/
│   ├── server.py               # FastMCP server + 4 tools
│   ├── source.py               # Source model (YAML loader)
│   ├── fetcher.py              # HTML → Markdown converter
│   ├── cache.py                # TTL cache manager
│   ├── discovery.py            # Sitemap-based page discovery
│   ├── search.py               # TF-IDF search engine
│   └── config.py               # Global configuration
└── tests/

许可证

MIT — 参见 LICENSE。

OpenGerwin MCP Servers 的一部分

Available Tools

4 tools
get_docA

Get full content of a specific documentation page.

Args: path: Documentation page path, e.g.: GEAP paths: - "models/gemini/3-1-pro" - "build/runtime/quickstart" - "scale/memory-bank/setup" - "govern/policies/overview" - "optimize/evaluation/agent-evaluation" - "agent-studio/overview" Vertex AI paths: - "multimodal/function-calling" - "rag-engine/rag-overview" - "models/gemini/2-5-flash" source: "geap" (default) or "vertex-ai"

Returns: Complete page content in Markdown format. If not cached, fetches live from the documentation site.

ParametersJSON Schema
NameRequiredDescriptionDefault
pathYes
sourceNogeap

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A3.7/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden. It mentions caching behavior and return format, but fails to disclose authentication requirements, rate limits, error handling, or what happens with invalid paths.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise and well-structured with a clear purpose sentence followed by Args/Returns. The examples are useful and do not feel excessive. Every sentence contributes to understanding.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple retrieval tool with two parameters and an output schema (implied), the description covers key aspects: purpose, parameters with examples, caching behavior, and return format. It is largely complete given the tool's complexity, though missing error state details.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 0%, so the description must compensate. It adds value by providing concrete path examples and explaining the source parameter's allowed values ('geap' or 'vertex-ai'). However, it doesn't clarify the exact format required for paths beyond examples.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states 'Get full content of a specific documentation page', using a specific verb and resource. It distinguishes from sibling tools like list_models, list_sections, and search_docs by focusing on a single page's full content.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides context by listing example paths for different sources and explaining the source parameter. It implicitly guides when to use this tool versus alternatives, but lacks explicit 'when not to use' instructions or direct comparisons.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

list_modelsA

List all available AI models on Google's platform.

Returns a quick reference of all models organized by family: Google (Gemini, Imagen, Veo, Lyria), Partners (Claude, Grok, Mistral, Llama), and Open Models (DeepSeek, Qwen, Kimi, etc.).

ParametersJSON Schema
NameRequiredDescriptionDefault

No parameters

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description must carry the burden. It describes the output as a 'quick reference' organized by family, but does not disclose potential behavioral aspects such as authentication requirements, rate limits, or whether it is read-only.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, front-loaded with the core purpose, and zero wasted words. Highly efficient.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given no parameters and an existing output schema, the description is fairly complete: it specifies what is listed and how it is organized. It could mention authentication or that it is a simple read operation, but overall adequate.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool has zero parameters, so schema coverage is effectively 100%. The description does not add parameter info, but none is needed. Baseline for 0 parameters is 4.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool lists all available AI models on Google's platform, organized by family. This specific verb+resource combination distinguishes it from sibling tools that deal with documents.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage for listing models, but does not explicitly state when to use this tool versus alternatives. There is no exclusion or guidance on prerequisites, though siblings are clearly document-focused.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

list_sectionsB

List all documentation sections and their page counts.

Args: source: "geap" (default) or "vertex-ai"

Returns: Structured overview of all available documentation sections with descriptions and page counts.

ParametersJSON Schema
NameRequiredDescriptionDefault
sourceNogeap

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

B3.3/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Description indicates it's a read operation returning structured data, but lacks details on side effects, auth requirements, or performance implications. Barely adequate given no annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Description is concise and well-structured with Args and Returns sections, but could be slightly more efficient.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Simple tool with one optional parameter and an output schema; description covers basic functionality but doesn't elaborate on scope or limitations.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Description explains the 'source' parameter with allowed values, adding value beyond the schema which lacks property descriptions.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

Description clearly states it lists documentation sections with page counts, distinguishing from sibling tools like get_doc and search_docs.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance on when to use this tool vs alternatives; no exclusions or prerequisites mentioned.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

search_docsA

Search Google AI platform documentation.

Args: query: Search terms (e.g. "function calling", "Memory Bank setup", "Agent Development Kit", "Gemini 3.1 Pro") max_results: Number of results to return (default: 5, max: 20) source: Documentation source: - "geap" (default) — Gemini Enterprise Agent Platform (current) - "vertex-ai" — Vertex AI Generative AI (legacy)

Returns: Matching documentation pages with titles, paths, and excerpts. Use get_doc(path) to read the full content of any result.

ParametersJSON Schema
NameRequiredDescriptionDefault
queryYes
max_resultsNo
sourceNogeap

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.7/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description details the return format (titles, paths, excerpts) and parameter behavior (defaults, max results, source options). It does not mention rate limits or authorization, but as a search tool with no annotations, the description provides sufficient transparency.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise and well-structured with bullet points for arguments. Every sentence adds value, and there is no redundant information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers purpose, parameters, return value, and sibling tools. It could mention result ordering or pagination, but overall it is comprehensive for a search tool with an output schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description thoroughly explains each parameter: query with examples, max_results with default and maximum, and source with options and defaults. Since schema coverage is 0%, the description fully compensates by providing clear semantics.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states 'Search Google AI platform documentation' and specifies it returns matching pages with titles, paths, and excerpts. It distinguishes from sibling get_doc which is for reading full content.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit guidance on when to use this tool (searching documentation) and recommends an alternative (get_doc for reading full content). It also differentiates between documentation sources (geap vs vertex-ai).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 4 tool updatesv0.1.0
    • First observedget_doc
    • First observedlist_models
    • First observedlist_sections
    • First observedsearch_docs
Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

A4.1/5.0

Scored across 4 tools

Disambiguation5/5

Each tool has a distinct purpose: get_doc retrieves page content, list_models lists models, list_sections lists documentation sections, and search_docs searches across docs. No ambiguity.

Naming Consistency5/5

All tool names follow a consistent verb_noun pattern (get_doc, list_models, list_sections, search_docs), making them predictable and easy to understand.

Tool Count5/5

With 4 tools, the server is well-scoped for a documentation access interface, covering essential operations without bloat.

Completeness5/5

The tool set provides comprehensive coverage for documentation: listing models, browsing sections, searching, and retrieving full content. No obvious gaps.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers