Oxylabs 远程 MCP 服务器 – 官方端点、工具列表与一键连接指南
Oxylabs 官方托管的远程 MCP 服务器。Oxylabs 远程 MCP,让代理通过高质量代理网络抓取任意网站,获取渲染后页面与结构化数据。
本页收录其官方远程端点、传输协议、认证方式、全部工具 Schema、实时可用性拨测与安全审计结果,并为 Cursor、Claude Code、VS Code、ChatGPT 等 10 大客户端提供一键接入指引。
Or add to ~/.cursor/mcp.json:
{
"mcpServers": {
"oxylabs": {
"url": "https://mcp.oxylabs.io/mcp"
}
}
}claude mcp add --transport http oxylabs https://mcp.oxylabs.io/mcp
Or add to .vscode/mcp.json:
{
"servers": {
"oxylabs": {
"type": "http",
"url": "https://mcp.oxylabs.io/mcp"
}
}
}- Open Settings → Connectors (Integrations) → Add Custom Connector.
- Paste the endpoint URL:
https://mcp.oxylabs.io/mcp - No authentication needed — tools are available immediately.
Bridge the remote endpoint into stdio with mcp-remote in claude_desktop_config.json:
{
"mcpServers": {
"oxylabs": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.oxylabs.io/mcp"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"oxylabs": {
"serverUrl": "https://mcp.oxylabs.io/mcp"
}
}
}Add to ~/.kiro/settings/mcp.json (or .kiro/settings/mcp.json in your workspace), then reopen Kiro's Agent panel to pick up the server:
{
"mcpServers": {
"oxylabs": {
"url": "https://mcp.oxylabs.io/mcp"
}
}
}Add to ~/.codex/config.toml — Codex reads remote HTTP servers from [mcp_servers.*] tables:
[mcp_servers.oxylabs] url = "https://mcp.oxylabs.io/mcp"
Add to cline_mcp_settings.json (Cline → MCP Servers → Configure), or paste it via Cline's MCP marketplace file:
{
"mcpServers": {
"oxylabs": {
"type": "streamableHttp",
"url": "https://mcp.oxylabs.io/mcp"
}
}
}Add to ~/.gemini/settings.json under mcpServers — httpUrl for Streamable HTTP, url for legacy SSE:
{
"mcpServers": {
"oxylabs": {
"httpUrl": "https://mcp.oxylabs.io/mcp"
}
}
}🧰 Oxylabs MCP 工具与 Schema(10)
| Tool | Safety | Description | Parameters |
|---|---|---|---|
| ai_crawler ~480 tok | 🟢 Read | Tool useful for crawling a website from starting url and returning data in a specified format. Schema is required only if output_format is json, csv or toon. 'render_javascript' is used to render javascript heavy websites. 'return_sources_limit' is used to limit the number of sources to return, for example if you expect results from single source, you can set it to 1. | url*: stringuser_prompt*: stringoutput_format: stringschemarender_javascript: booleanreturn_sources_limit: integergeo_location |
| ai_scraper ~375 tok | 🟢 Read | Scrape the contents of the web page and return the data in the specified format. Schema is required only if output_format is json or csv. 'render_javascript' is used to render javascript heavy websites. | url*: stringoutput_format: stringschemarender_javascript: booleangeo_location |
| ai_browser_agent ~356 tok | 🟢 Read | Run the browser agent and return the data in the specified format. This tool is useful if you need navigate around the website and do some actions. It allows navigating to any url, clicking on links, filling forms, scrolling, etc. Finally it returns the data in the specified format. Schema is required only if output_format is json, csv or toon. 'task_prompt' describes what browser agent should achieve | url*: stringtask_prompt*: stringoutput_format: stringschemageo_location |
| ai_search ~300 tok | 🟢 Read | Search the web based on a provided query. 'return_content' is used to return markdown content for each search result. If 'return_content' is set to True, you don't need to use ai_scraper to get the content of the search results urls, because it is already included in the search results. if 'return_content' is set to True, prefer lower 'limit' to reduce payload size. | query*: stringlimit: integerrender_javascript: booleanreturn_content: booleangeo_location |
| generate_schema ~78 tok | 🟢 Read | Generate a json schema in openapi format. | user_prompt*: stringapp_name*: string |
| ai_map ~405 tok | 🟢 Read | Tool useful for mapping website's URLs. | url*: stringsearch_keywordsuser_promptmax_crawl_depth: integerrender_javascript: booleanlimit: integergeo_locationallow_subdomains: booleanallow_external_domains: boolean |
| universal_scraper ~511 tok | 🟢 Read | Get a content of any webpage. Supports browser rendering, parsing of certain webpages and different output formats. | url*: stringrenderuser_agent_typegeo_locationoutput_format |
| google_search_scraper ~848 tok | 🟢 Read | Scrape Google Search results. Supports content parsing, different user agent types, pagination, domain, geolocation, locale parameters and different output formats. | query*: stringparse: booleanrenderuser_agent_typestart_page: integerpages: integerlimit: integerdomaingeo_locationlocalead_mode: booleanoutput_format |
| amazon_search_scraper ~924 tok | 🟢 Read | Scrape Amazon search results. Supports content parsing, different user agent types, pagination, domain, geolocation, locale parameters and different output formats. Supports Amazon specific parameters such as category id, merchant id, currency. | query*: stringcategory_idmerchant_idcurrencyparse: booleanrenderuser_agent_typestart_page: integerpages: integerdomaingeo_locationlocaleoutput_format |
| amazon_product_scraper ~849 tok | 🟢 Read | Scrape Amazon products. Supports content parsing, different user agent types, domain, geolocation, locale parameters and different output formats. Supports Amazon specific parameters such as currency and getting more accurate pricing data with auto select variant. | query*: stringautoselect_variant: booleancurrencyparse: booleanrenderuser_agent_typedomaingeo_locationlocaleoutput_format |
🛡️ 安全评级、质量与上下文 Token 审计
- No prompt-injection or hidden-character anomalies detected in tool descriptions
- Streamable HTTP primary endpoint (current 2025-03-26 transport)
- Public read-only-friendly endpoint — no credentials stored client-side
- Official first-party endpoint operated by the vendor
- Manually verified by RemoteMCP.net
🔐 OAuth 2.1 与安全
- 认证方式
- 免密公开
- 动态客户端注册(DCR)
- —
- PKCE S256
- —
- 受保护资源元数据
- https://mcp.oxylabs.io/.well-known/oauth-protected-resource
📡 官方远程端点与传输协议
远程端点 https://mcp.oxylabs.io/mcp 采用 Streamable HTTP(流式 HTTP,2025-03-26 规范)。Streamable HTTP 在单一 POST 端点上完成 JSON-RPC 双向交互,是当前 MCP 规范推荐的传输方式,可直接被 Claude、Cursor、VS Code 等客户端以 URL 形式接入,无需 npx mcp-remote 桥接。
接入前可将端点粘贴至本站 在线 MCP Inspector 实测 initialize / tools/list 握手,确认真实协议行为。
🔐 认证与权限机制
认证方式为 免密公开(None)。完全免密公开,直接粘贴 URL 即可使用——适合文档检索与公开数据查询类场景。
🛡️ 上下文开销与安全审计
该服务端当前登记 10 个工具,全部定义合计约 5146 tokens,会在每次请求时注入模型系统提示词。安全评级为 A+,官方直营来源。工具名与 Schema 越精简,对上下文窗口的"税"越低;可用 feecalculator.app 估算每次调用的边际成本。
🧩 主流客户端快速接入
本站为 Oxylabs 生成了 Cursor / Claude Code / VS Code / ChatGPT / Windsurf / Kiro / Codex / Cline / Gemini 等 10 大客户端的专属配置——包括一键 deeplink、claude mcp add 命令与 JSON/TOML 片段。下方选择你的客户端即可查看 1 分钟接入步骤。代表性工具包括 ai_crawler、ai_scraper、ai_browser_agent、ai_search、generate_schema 等。
📖 Oxylabs 详解
Oxylabs — Remote MCP Server
Oxylabs' remote MCP lets agents scrape any website through premium proxy infrastructure — fetch rendered pages, bypass geo-restrictions and extract clean structured data. The remote endpoint means there is nothing to install — point your MCP client at the URL and start calling tools right away. Oxylabs operates and versions the server themselves, so tool coverage tracks their API surface as it ships.
Official Oxylabs Endpoint & Transport Architecture
- Endpoint URL:
https://mcp.oxylabs.io/mcp - Transport: Streamable HTTP
Streamable HTTP is the current MCP transport (2025-03-26 spec): a single URL accepts JSON-RPC POSTs and may stream responses over SSE, so one endpoint serves both simple and streaming workloads. Oxylabs does not publish a separate legacy SSE endpoint, so your client must speak Streamable HTTP.
Authentication, Scopes & Permission Boundaries
Oxylabs is public and unauthenticated — no OAuth, no API key, no headers. Any MCP client can connect and call its tools directly, which also makes it a great endpoint for smoke-testing your own MCP client or proxy.
All 0 cataloged tools are read-only — Oxylabs cannot mutate your account data, which makes it safe to connect broadly.
Common Troubleshooting & Connection Tips
- "Session expired" / stale
Mcp-Session-Id— just re-sendinitialize; stateless clients can ignore session headers entirely. - Production scraping volume requires an Oxylabs account and credentials.
🚀 如何接入 Oxylabs 远程 MCP
- 获取远程端点 — 复制官方端点 https://mcp.oxylabs.io/mcp。
- 写入客户端配置 — 在 Cursor 的 mcp.json / Claude Code CLI / 其他客户端中添加该 URL(见下方分客户端指南)。
- 授权并验证 — 保存配置后,在本站 Inspector 中运行 initialize 与 tools/list 验证连通性。
🔌 Oxylabs × 各客户端连接指南
❓ 常见问题
Oxylabs 远程 MCP 服务器是官方提供的吗?
本条目标记为「官方直营」。端点由 Oxylabs 官方运营,所有收录端点均经过真实 MCP initialize 握手验证。
如何在 Claude / Cursor 中连接 Oxylabs?
复制远程端点 https://mcp.oxylabs.io/mcp:Claude Code 执行 claude mcp add --transport http oxylabs https://mcp.oxylabs.io/mcp;Cursor 可直接使用本页的一键 deeplink 或在 ~/.cursor/mcp.json 中添加 { "mcpServers": { "oxylabs": { "url": "https://mcp.oxylabs.io/mcp" } } }。
连接时返回 401 或需要密钥怎么办?
该端点免密公开;若仍返回 401 请确认 URL 未拼写错误,并在 Inspector 中重试。
本地 stdio MCP 和这个远程 MCP 有什么区别?
本地 MCP 通过 npx/uvx 在你的机器上启动 stdio 子进程,依赖本地运行时且密钥以明文写入配置文件;远程 MCP 是 Oxylabs 托管的 HTTPS 端点,免安装、支持 OAuth 2.1 授权流,且同一端点可跨桌面与移动端复用。详见 本地 vs 远程 MCP 对比。