Browser Use
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
作者提供的说明(英文原文)
Web scraping, extraction and search for AI agents. Hosted MCP with workspace API keys.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
Authorization你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
在终端中运行这条命令
claude mcp add --transport http deepsfetch https://mcp.deepsfetch.com/mcp --header 'Authorization: Bearer {api_key}'运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"deepsfetch": {
"url": "https://mcp.deepsfetch.com/mcp",
"headers": {
"Authorization": "Bearer {api_key}"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"deepsfetch": {
"type": "http",
"url": "https://mcp.deepsfetch.com/mcp",
"headers": {
"Authorization": "Bearer {api_key}"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
你的选择通过 Cookie 保存在这台设备上,无需账户。
Authorization HTTP header这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月6日
注册名称已通过 GitHub 账户 ermultimidia 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(2.0.1)于26天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Browser Use, Chrome DevTools MCP, MCP Server Browser
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
MCP server for Chrome DevTools
MCP server for browser use access
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
English | Português
⭐ If these n8n workflows, SDKs, or MCP tools save you time, please give this repository a Star on GitHub! It helps other automation builders discover the toolkit.
[Complete manual: 33 step-by-step guides](docs/manual/README.en.md) — API, MCP, SDKs, CLI, collections, monitors, files, integrations and billing.
Collect web content for AI agents and automation with explicit credit limits.
Open-source JavaScript and Python clients, a CLI, and n8n integration code for the hosted DeepSFetch API. Start with one public HTML page, inspect the result, and build from there.
This repository contains client-side tools. The collection engine and hosted service are not included. Running the examples requires a DeepSFetch account and API key; collection consumes workspace credits. See current plans.
Automate structured HTML extraction in n8n with explicit credit caps, idempotency keys, and field validation before sending data to your alerts, spreadsheets, or AI/RAG pipelines.
| Ready Template | Use Case | What It Extracts & Delivers | Setup & Download |
|---|---|---|---|
| Price Monitoring | Track product price changes on a schedule | Validated name, price, and currency (JSON-LD / CSS) with baseline comparison | Configure & Download JSON · n8n Guide |
| Product Catalog | Extract up to 100 items from a listing page | Structured item arrays with per-item field validation (no pagination) | Configure & Download JSON · n8n Guide |
| AI / RAG Documentation | Feed clean documentation into vector stores or LLMs | Source metadata, stable document IDs, content hash (change detection), and chunks | Configure & Download JSON · n8n Guide |
n8n-nodes-deepsfetch)For self-hosted n8n, install via the UI (Settings → Community Nodes → Install → n8n-nodes-deepsfetch) or manually in your community nodes folder:
npm install n8n-nodes-deepsfetch
*(Using n8n Cloud? You can download the standard HTTP Request workflow templates directly from Guided Setup without installing a community node.)*
📸 View n8n Node Configuration & Validated JSON Output Screenshots
1. Starting a Collection (`Start Collection` operation with payload & language selection):
2. Validated Output & Price Comparison (`Compare price` step with structured JSON result):
Requires Node.js 22+. Create a key in API keys. From the repository root, set it for the current terminal session and run the dependency-free example:
PowerShell:
$env:DEEPSFETCH_API_KEY = 'YOUR_DEEPSFETCH_API_KEY'
node examples/scrape.mjs https://example.com
macOS / Linux:
export DEEPSFETCH_API_KEY='YOUR_DEEPSFETCH_API_KEY'
node examples/scrape.mjs https://example.com
The example submits one HTTP-only collection with a maximum of 3 DeepSFetch credits and browser fallback disabled, then polls the same job. It prints the completed job as JSON. This cap is not a promise of a free or successful collection. JavaScript-dependent pages need another configuration. See getting started for failure handling and resuming a job.
| Tool | Install or guide | What is included |
|---|---|---|
| JavaScript SDK | npm install @deepsfetch/sdk | API v2 client and job polling |
| CLI | npm install -g @deepsfetch/cli | Commands for collection, jobs and estimates |
| Python SDK | python -m pip install ./packages/deepsfetch-sdk-python | Source installation; PyPI release pending |
| MCP | Remote connection guide | Configuration for the hosted MCP endpoint |
| n8n | npm install n8n-nodes-deepsfetch · Guide | Source for the npm-published community node |
The remote MCP server implementation is not included here. The n8n node is not verified by n8n. The remote MCP is listed in the official registry. Native partnerships and a self-hosted collection engine are not claimed by this repository.
After installing the CLI and setting the environment variable above:
deepsfetch scrape --file examples/scrape-request.json --wait
The CLI writes its idempotency key to stderr. If a submission's outcome is unknown, retry the same request with --idempotency-key THE_SAME_KEY. Once you have a job ID, use deepsfetch job JOB_ID --wait instead of submitting another collection.
| Capability | In this repository | Hosted DeepSFetch |
|---|---|---|
| JavaScript SDK, Python SDK, CLI | Source and tests | Connect to API v2 |
| n8n community node | Source distributed as JavaScript | Calls collection integration endpoints |
| MCP | Connection guide | Remote server |
| Collection engine, browser infrastructure | Not included | Managed by DeepSFetch |
| Accounts, credits, billing and dashboard | Not included | Managed by DeepSFetch |
npm install --ignore-scripts
npm test npm run check
Run Python tests from the repository root:
python -m pip install ./packages/deepsfetch-sdk-python
python -m unittest discover -s packages/deepsfetch-sdk-python/tests
Tests use fixtures and a local HTTP server; they do not submit paid collections. See contributing, security, and the roadmap.
MIT license applies to the tools in this repository. Using the hosted service is subject to its own terms and credit limits.
Read the first-collection tutorial, run the operational smoke test, and share your first-run experience.
The authenticated Firecrawl comparison includes results, limitations and reproducible code. The published DEV tutorial walks through signup, first collection and MCP. Also see the Smithery listing, with instructions and limits in the MCP guide.