AUDISTO-MCP

MCP浏览器工具

提供方 tentaclequing

浏览器工具 stdio · 本地进程 API Key 社区 / 第三方 Claude DesktopClaudeGemini CLIZed
查看接入方式

AUDISTO-MCP is a read-only MCP server that connects AI agents to Audisto's technical SEO analysis platform, providing tools to list recent crawls and retrieve detailed crawl summaries, enabling AI clients like Claude and Gemini to query website crawl data.

一句话选型结论

按资料字段整理

AUDISTO-MCP 属于浏览器工具类 MCP 服务器,stdio · 本地进程,API Key,可直接接入 Claude Desktop、Claude、Gemini CLI 等,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类浏览器工具 传输方式stdio · 本地进程 鉴权API Key 部署方式本地部署
GitHub 星标 2 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

浏览器工具
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

Local Deployment MCP server Technical SEO Python Crawl data AI agent integration Website analysis Read-only access

支持客户端

适用场景

Local DeploymentMCP serverTechnical SEO

安装与配置

stdio · 本地进程 API Key
客户端在本机拉起子进程,数据不出本机
python C:\Users\YourUsername\Documents\GitHub\AUDISTO-MCP\server.py

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 python,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
自动操作网页

打开页面、点击、填表,把浏览器动作交给模型执行。

抓取结构化数据

把页面内容整理成表格或 JSON,省掉手工复制粘贴。

关键流程回归

按步骤检查页面是否正常,回归时少漏环节。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Claude、Gemini CLI,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 使用前要准备好API Key对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在浏览器工具场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
AUDISTO-MCP当前 未标注 stdio · 本地进程 本地部署 有 2026-02-07
Playwright MCP Server 21 stdio · 本地进程 本地部署 有 2026-05-18
Fetch 9a2682de 1 SSE · 服务端推送 支持远程部署 有 2026-05-18
Browser 738d52bc 未标注 stdio · 本地进程 本地部署 有 2026-05-17
MCP 40600ece 未标注 stdio · 本地进程 本地部署 有 2026-05-17
Skyvern MCP 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

浏览器工具
字段来自项目仓库与公开索引,中文说明由本站标注
提供方tentaclequing上游标注的维护方 tentaclequing
分类浏览器工具本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式API Key远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · python本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-02-07上游最后核验时间

详细介绍

8.3 KB
来自项目仓库与上游目录的原文,未做删改

AUDISTO-MCP

An Audisto MCP (Model Context Protocol) server for connecting AI agents to Audisto's technical SEO analysis platform.

What is this?

AUDISTO-MCP is a bridge that lets AI agents (like Claude, via Claude Desktop) access your Audisto crawl data through a standardized interface. It exposes tools to list recent crawls and retrieve detailed crawl summaries.

Read-only: This MCP server is intentionally read-only — it does NOT expose any endpoints to start, stop, or modify Audisto crawls. It only retrieves and reports data from Audisto's API.

API Usage Limits

Audisto's API enforces strict usage limits:
- One request per API key at a time — If a second request is sent while one is still executing, you'll receive a 429 (Too Many Requests) error. The MCP client includes automatic retries with backoff to handle this gracefully.
- Chunk size limit — Maximum 10,000 items per request. The client validates this automatically.
- Important: Only run one MCP instance per API key at a time. If you run Claude Desktop and Gemini CLI simultaneously with the same credentials, they may conflict and cause 429 errors.

Quick Start

Prerequisites

  • Python 3.8 or later
  • An Audisto account with API credentials

Installation

1. Clone this repository```bash

git clone https://github.com/tentaclequing/AUDISTO-MCP.git
cd AUDISTO-MCP


#### 2. Create a virtual environment

**Windows (PowerShell):**```powershell
python -m venv .venv
.\.venv\Scripts\Activate.ps1

Linux / macOS (Bash/Zsh):```bash
python3 -m venv .venv
source .venv/bin/activate


#### 3. Install dependencies```bash
pip install -r requirements.txt

4. Set up credentials

Copy the example environment file and add your credentials:```bash
cp .env.example .env


Then edit `.env` with your Audisto API credentials:```
AUDISTO_API_KEY=your_api_key_here
AUDISTO_PASSWORD=your_password_here

Get your API credentials from:

  • Web:
  • Documentation:

Windows (PowerShell): Load environment variables```powershell
Get-Content .env | ForEach-Object {
if ($_ -match '^([^=]+)=(.*)$') {
Environment::SetEnvironmentVariable($matches[1], $matches[2])
}
}


**Linux / macOS (Bash/Zsh):** Load environment variables```bash
export $(cat .env | xargs)

Or simply export them manually:```bash
export AUDISTO_API_KEY=your_api_key_here
export AUDISTO_PASSWORD=your_password_here


#### 5. Run the server

**Windows (PowerShell):**```powershell
python server.py

Linux / macOS (Bash/Zsh):```bash
python3 server.py


You should see:```
INFO - Validating startup credentials...
INFO - Credentials validated successfully
INFO - MCP Server initialized

Available Tools

get_help()

Displays all available commands, usage examples, and workflow tips.
- Use this if you're unsure what to ask
- Example: "Help" or "What can I do with this?"

get_crawl_status()

Lists the last 5 crawls from your Audisto account with:
- Crawl ID
- Domain
- Status (finished or in progress)

get_crawl_summary(crawl_id)

Retrieves details about a specific crawl:
- Pages crawled
- Max depth reached
- Start time
- Domain info

Supported AI Clients

AUDISTO-MCP works with any MCP-compatible AI client. Here's how to set it up for each major platform:

Claude Desktop

The easiest way to use Claude with MCP servers on your desktop.

  1. Find your claude_desktop_config.json:
    - Windows: %APPDATA%\Claude\claude_desktop_config.json
    - macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    - Linux: ~/.config/Claude/claude_desktop_config.json

  2. Add this MCP server to the mcpServers section:

Windows (use backslashes or forward slashes):json { "mcpServers": { "audisto": { "command": "python", "args": ["C:\\Users\\YourUsername\\Documents\\GitHub\\AUDISTO-MCP\\server.py"], "env": { "AUDISTO_API_KEY": "your_api_key_here", "AUDISTO_PASSWORD": "your_password_here" } } } }

macOS / Linux:json { "mcpServers": { "audisto": { "command": "python3", "args": ["/Users/YourUsername/path/to/AUDISTO-MCP/server.py"], "env": { "AUDISTO_API_KEY": "your_api_key_here", "AUDISTO_PASSWORD": "your_password_here" } } } }

  1. Restart Claude Desktop
  2. Ask Claude questions like: "What crawls do I have?" or "Show me crawl 12345"

Claude CLI

Use Claude directly from your terminal.

Prerequisites

  • Claude CLI installed: https://github.com/anthropics/claude-cli

Setup

Windows (PowerShell):```powershell

Set the path to your server

$serverPath = "C:\Users\YourUsername\Documents\GitHub\AUDISTO-MCP\server.py"

Add MCP server

claude mcp add audisto "python $serverPath"

Verify

claude mcp list


**macOS / Linux (Bash/Zsh):**```bash
# Set the path to your server
SERVER_PATH="$HOME/path/to/AUDISTO-MCP/server.py"

# Add MCP server
claude mcp add audisto "python3 $SERVER_PATH"

# Verify
claude mcp list

Set credentials

Windows (PowerShell):```powershell


**macOS / Linux (Bash/Zsh):**```bash
export AUDISTO_API_KEY="your_api_key_here"
export AUDISTO_PASSWORD="your_password_here"

Usage```bash

claude "What crawls do I have?"
claude "Show me crawl 12345"


### Gemini CLI
Use Gemini directly from your terminal.

#### Prerequisites
- Gemini CLI installed: `npm install -g @google/generative-ai-cli` or follow official docs

#### Setup (all platforms)

First, make sure your virtual environment is activated and Python can find the server:

**Windows (PowerShell):**```powershell
# Store full path to server.py
$serverPath = "C:\Users\YourUsername\Documents\GitHub\AUDISTO-MCP\server.py"

# Add MCP server
gemini mcp add audisto "python $serverPath" --scope user

# Enable it
gemini mcp enable audisto

# Verify
gemini mcp list

macOS / Linux (Bash/Zsh):```bash

Store full path to server.py

SERVER_PATH="$HOME/path/to/AUDISTO-MCP/server.py"

Add MCP server

gemini mcp add audisto "python3 $SERVER_PATH" --scope user

Enable it

gemini mcp enable audisto

Verify

gemini mcp list


### Set credentials

After adding the server, set environment variables:

**Windows (PowerShell):**```powershell
[Environment]::SetEnvironmentVariable("AUDISTO_API_KEY", "your_api_key_here", "User")
[Environment]::SetEnvironmentVariable("AUDISTO_PASSWORD", "your_password_here", "User")

macOS / Linux (Bash/Zsh):```bash

Add to ~/.bashrc, ~/.zshrc, or ~/.profile

export AUDISTO_API_KEY="your_api_key_here"
export AUDISTO_PASSWORD="your_password_here"

Then reload

source ~/.bashrc # or ~/.zshrc


### Remove the server (if needed)

**All platforms:**```bash
gemini mcp remove audisto

Asking Questions

You can now ask your AI agent questions like:
- "What crawls do I have in Audisto?"
- "Show me the summary for crawl 12345"
- "How many pages were crawled in my latest crawl?"
- "List all my Audisto crawls and their statuses"

Security checklist

Follow these before publishing or sharing the repository:

  • Protect credentials: Do not commit API keys or passwords. Use environment variables or a secrets manager.
  • .env handling: If you use a .env file for local testing, add it to .gitignore and never commit it.
  • Supply-chain scan: Run pip-audit or safety before releasing to check dependency vulnerabilities.
  • CI checks: Add a CI job to run pip-audit and a linter (e.g., flake8) on PRs.
  • Least privilege: Use an API key with only the necessary Audisto scopes.
  • Credential rotation: Rotate API keys if they were accidentally exposed.
  • Limited exposure: Prefer launching the MCP server via stdio (CLI integrations) instead of binding network sockets; if network binding is required, restrict to localhost and/or use firewall rules.
  • Logs: Avoid logging sensitive data (API keys, full request/response bodies). Keep logs for diagnostics only.

This repository includes a GitHub Actions CI workflow that automatically runs pip-audit, code linting (ruff), type checking (mypy), and tests on all pushes and pull requests.

License

MIT - see LICENSE

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
AUDISTO-MCP 怎么安装?

在客户端按下面的安装命令接入:python C:\Users\YourUsername\Documents\GitHub\AUDISTO-MCP\server.py。装完在客户端配置里引用同一个包或端点。

AUDISTO-MCP 怎么在 Claude Desktop、Claude、Gemini CLI 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

AUDISTO-MCP 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

AUDISTO-MCP 需要 API Key 吗?

资料标注需要API Key认证,使用前准备好对应凭据,并按最小权限配置。

AUDISTO-MCP 支持哪些客户端?

资料标注支持 Claude Desktop、Claude、Gemini CLI、Zed;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

AUDISTO-MCP 能提供哪些能力?

资料显示能力标签包括 Local Deployment、MCP server、Technical SEO、Python、Crawl data、AI agent integration。具体工具清单以官方文档为准。

AUDISTO-MCP 需要什么运行环境?

运行时为 python;属于本地进程,需要在本机装好运行环境。

AUDISTO-MCP 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-02-07。