Skip to main content
Glama
actors-mcp-server

actors-mcp-server

Official
by apify

Apify 模型上下文协议 (MCP) 服务器

Actors MCP 服务器 铁匠徽章

为所有Apify Actors实现一个 MCP 服务器。该服务器支持与一个或多个 Apify Actors 进行交互,这些 Actors 可以在 MCP 服务器配置中定义。

该服务器可以以两种方式使用:

  • 🇦 MCP 服务器 Actor – 可通过服务器发送事件 (SSE) 访问的 HTTP 服务器,请参阅指南

  • ⾕ MCP Server Stdio – 可通过标准输入/输出 (stdio) 访问的本地服务器,请参阅指南

您还可以使用类似聊天的 UI 与 MCP 服务器进行交互,使用 💬 Tester MCP Client

🎯 Apify MCP 服务器做什么?

MCP 服务器 Actor 允许 AI 助手使用任意Apify Actor作为工具来执行特定任务。例如,它可以:

MCP 客户端

要与 Apify MCP 服务器交互,您可以使用 MCP 客户端,例如:

当您将 Actors 与 MCP 服务器集成时,您可以询问:

  • “搜索网络并总结有关人工智能代理的最新趋势”

  • “寻找旧金山十大最佳意大利餐厅”

  • “查找并分析巨石强森的 Instagram 个人资料”

  • “提供使用带有源 URL 的模型上下文协议的分步指南”

  • “我可以使用哪些 Apify Actors?”

下图展示了 Apify MCP 服务器与 Apify 平台和 AI 客户端的交互方式:

Actors-MCP-服务器

使用 MCP Tester 客户端,您可以动态加载 Actor,但其他 MCP 客户端尚不支持此功能。我们还计划添加更多功能,详情请参阅路线图。

🔄 什么是模型上下文协议?

模型上下文协议 (MCP) 允许 AI 应用程序(以及 AI 代理)(例如 Claude Desktop)连接到外部工具和数据源。MCP 是一种开放协议,支持 AI 应用程序、AI 代理以及本地或远程资源之间安全、可控的交互。

有关更多信息,请参阅模型上下文协议网站或博客文章什么是 MCP 以及它为什么重要? 。

🤖 MCP 服务器与 AI 代理有何关系?

Apify MCP 服务器通过 MCP 协议公开 Apify 的 Actors,允许实现 MCP 协议的 AI Agent 或框架访问所有 Apify Actors,作为数据提取、网络搜索和其他任务的工具。

要了解更多关于 AI 代理的信息,请阅读我们的博客文章:什么是 AI 代理?并浏览 Apify 精选的AI 代理合集。有兴趣在 Apify 上构建自己的 AI 代理并实现盈利吗?查看我们的分步指南,了解如何在 Apify 平台上创建、发布和实现 AI 代理盈利。

🧱 组件

工具

演员

任何Apify Actor均可用作工具。默认情况下,服务器已预先配置了以下指定的 Actor,但可以通过提供 Actor 输入来覆盖此配置。

'apify/instagram-scraper'
'apify/rag-web-browser'
'lukaskrivka/google-maps-with-contact-details'

MCP 服务器加载 Actor 输入架构,并创建与 Actor 对应的 MCP 工具。请参阅RAG Web 浏览器的输入架构示例 。

工具名称必须始终是完整的 Actor 名称,例如apify/rag-web-browser 。MCP 工具的参数表示 Actor 的输入参数。例如,对于apify/rag-web-browser Actor,其参数为:

{
  "query": "restaurants in San Francisco",
  "maxResults": 3
}

您无需指定输入参数或调用哪个 Actor;所有操作均由 LLM 管理。调用工具时,LLM 会自动将参数传递给 Actor。您可以参考特定 Actor 的文档,了解可用参数的列表。

辅助工具

服务器提供了一组辅助工具来发现可用的参与者并检索其详细信息:

  • get-actor-details :检索有关特定 Actor 的文档、输入模式和详细信息。

  • discover-actors :使用关键字搜索相关的演员并返回他们的详细信息。

还有一些工具可以管理可用的工具列表。但是,动态添加和删除工具需要 MCP 客户端具备更新工具列表(处理ToolListChangedNotificationSchema )的能力,而这通常不受支持。

您可以使用Apify Tester MCP 客户端Actor 尝试此功能。要启用此功能,请设置enableAddingActors参数。

  • add-actor-as-tool :按名称将 Actor 添加到可用工具列表中但不执行它,需要用户同意稍后运行。

  • remove-actor-from-tool :当不再需要某个 Actor 时,从可用工具列表中按名称将其删除。

Related MCP server: Apify MCP Server

提示和资源

服务器不提供任何资源和提示。我们计划在未来添加Apify 的数据集和键值存储作为资源。

⚙️ 使用方法

Apify MCP 服务器可以通过两种方式使用:作为在 Apify 平台上运行的 Apify Actor或作为在您的机器上运行的本地服务器。

🇦 MCP 服务器 Actor

备用 Web 服务器

Actor 以待机模式运行,并带有一个接收和处理请求的 HTTP Web 服务器。

要使用默认 Actors 启动服务器,请将带有Apify API 令牌的HTTP GET 请求发送到以下 URL:

https://actors-mcp-server.apify.actor?token=<APIFY_TOKEN>

您也可以使用不同的 Actor 集合来启动 MCP 服务器。为此,请创建一个任务并指定要使用的 Actor 列表。

然后,使用选定的 Actors 在待机模式下运行任务:

https://USERNAME--actors-mcp-server-task.apify.actor?token=<APIFY_TOKEN>

您可以在Apify Store中找到所有可用演员的列表。

💬 通过 SSE 与 MCP 服务器交互

服务器运行后,您可以与服务器发送事件 (SSE) 进行交互,向服务器发送消息并接收响应。最简单的方法是使用 Apify 上的Tester MCP 客户端。

Claude Desktop目前不支持 SSE,但您可以将其与 Stdio 传输一起使用;有关更多详细信息,请参阅本地主机上的 MCP 服务器。注意:Claude Desktop 的免费版本可能会遇到与服务器的间歇性连接问题。

在客户端设置中,需要提供服务器配置:

{
    "mcpServers": {
        "apify": {
            "type": "sse",
            "url": "https://actors-mcp-server.apify.actor/sse",
            "env": {
                "APIFY_TOKEN": "your-apify-token"
            }
        }
    }
}

或者,您可以使用clientSse.ts脚本或使用curl </> 命令测试服务器。

  1. 通过向以下 URL 发送 GET 请求来启动服务器发送事件 (SSE):

    curl https://actors-mcp-server.apify.actor/sse?token=<APIFY_TOKEN>

    服务器将使用sessionId进行响应,您可以使用该 sessionId 向服务器发送消息:

    event: endpoint
    data: /message?sessionId=a1b
  2. 通过使用sessionId发出 POST 请求向服务器发送消息:

    curl -X POST "https://actors-mcp-server.apify.actor/message?token=<APIFY_TOKEN>&session_id=a1b" -H "Content-Type: application/json" -d '{
      "jsonrpc": "2.0",
      "id": 1,
      "method": "tools/call",
      "params": {
        "arguments": { "searchStringsArray": ["restaurants in San Francisco"], "maxCrawledPlacesPerSearch": 3 },
        "name": "lukaskrivka/google-maps-with-contact-details"
      }
    }'

    MCP 服务器将使用提供的参数作为输入参数,启动 Actor lukaskrivka/google-maps-with-contact-details 。对于此 POST 请求,服务器将响应:

    Accepted
  3. 接收响应。服务器将使用提供的查询参数调用指定的 Actor 作为工具,并通过 SSE 将响应流式传输回客户端。响应将以 JSON 文本形式返回。

    event: message
    data: {"result":{"content":[{"type":"text","text":"{\"searchString\":\"restaurants in San Francisco\",\"rank\":1,\"title\":\"Gary Danko\",\"description\":\"Renowned chef Gary Danko's fixed-price menus of American cuisine ... \",\"price\":\"$100+\"...}}]}}

本地主机上的 MCP 服务器

您可以通过 Claude Desktop 或任何其他MCP 客户端在本地计算机上配置 Apify MCP 服务器来运行它。您也可以使用Smithery自动安装服务器。

先决条件

  • MacOS 或 Windows

  • 必须安装最新版本的 Claude Desktop(或其他 MCP 客户端)

  • Node.js (v18 或更高版本)

  • Apify API 令牌( APIFY_TOKEN )

确保已正确安装node和npx :

node -v
npx -v

如果没有,请按照本指南安装 Node.js:下载并安装 Node.js 和 npm 。

克劳德桌面

要配置 Claude Desktop 与 MCP 服务器配合使用,请按照以下步骤操作。有关详细指南,请参阅Claude Desktop 用户指南或观看视频教程。

  1. 下载 Claude 桌面版

    • 适用于 Windows 和 macOS。

    • 对于 Linux 用户,您可以使用这个非官方的构建脚本来构建 Debian 包。

  2. 打开 Claude 桌面应用程序并从左上角的菜单栏启用开发者模式。

  3. 启用后,打开**“设置”** (也可以从左上角的菜单栏中打开)并导航到“开发人员选项” ,您将在其中找到**“编辑配置”**按钮。

  4. 打开配置文件并编辑以下文件:

    • 在 macOS 上: ~/Library/Application\ Support/Claude/claude_desktop_config.json

    • 在 Windows 上: %APPDATA%/Claude/claude_desktop_config.json

    • 在 Linux 上: ~/.config/Claude/claude_desktop_config.json

    {
     "mcpServers": {
       "actors-mcp-server": {
         "command": "npx",
         "args": ["-y", "@apify/actors-mcp-server"],
         "env": {
            "APIFY_TOKEN": "your-apify-token"
         }
       }
     }
    }

    或者,您可以使用actors参数来选择一个或多个 Apify Actors:

    {
    "mcpServers": {
      "actors-mcp-server": {
        "command": "npx",
        "args": [
          "-y", "@apify/actors-mcp-server",
          "--actors", "lukaskrivka/google-maps-with-contact-details,apify/instagram-scraper"
        ],
        "env": {
           "APIFY_TOKEN": "your-apify-token"
        }
      }
    }
    }
  5. 重启Claude桌面

    • 完全退出 Claude Desktop(确保它不仅仅是最小化或关闭)。

    • 重新启动 Claude Desktop。

    • 查找🔌图标以确认 Actors MCP 服务器已连接。

  6. 打开 Claude 桌面聊天并询问“我可以使用哪些 Apify Actors?”

    Claude 桌面与 Actors-MCP 服务器

  7. 示例

    您可以要求克劳德执行任务,例如:

    Find and analyze recent research papers about LLMs.
    Find the top 10 best Italian restaurants in San Francisco.
    Find and analyze the Instagram profile of The Rock.

VS 代码

对于一键安装,请单击以下安装按钮之一:

在 VS Code 中使用 NPX 安装 在 VS Code Insiders 中使用 NPX 安装

手动安装

您可以在 VS Code 中手动安装 Apify MCP 服务器。首先,点击本节顶部的安装按钮之一,即可一键安装。

或者,在 VS Code 中,将以下 JSON 块添加到“用户设置 (JSON)”文件中。您可以按下Ctrl + Shift + P并输入Preferences: Open User Settings (JSON)来执行此操作。

{
  "mcp": {
    "inputs": [
      {
        "type": "promptString",
        "id": "apify_token",
        "description": "Apify API Token",
        "password": true
      }
    ],
    "servers": {
      "actors-mcp-server": {
        "command": "npx",
        "args": ["-y", "@apify/actors-mcp-server"],
        "env": {
          "APIFY_TOKEN": "${input:apify_token}"
        }
      }
    }
  }
}

或者,您可以将其添加到工作区中名为.vscode/mcp.json的文件中 - 只需省略顶层mcp {}键即可。这样您就可以与其他人共享配置。

如果要指定加载哪些 Actors,可以添加--actors参数:

{
  "servers": {
    "actors-mcp-server": {
      "command": "npx",
      "args": [
        "-y", "@apify/actors-mcp-server",
        "--actors", "lukaskrivka/google-maps-with-contact-details,apify/instagram-scraper"
      ],
      "env": {
        "APIFY_TOKEN": "${input:apify_token}"
      }
    }
  }
}

VS 代码

对于一键安装,请单击以下安装按钮之一:

在 VS Code 中使用 NPX 安装 在 VS Code Insiders 中使用 NPX 安装

手动安装

您可以在 VS Code 中手动安装 Apify MCP 服务器。首先,点击本节顶部的安装按钮之一,即可一键安装。

或者,在 VS Code 中,将以下 JSON 块添加到“用户设置 (JSON)”文件中。您可以按下Ctrl + Shift + P并输入Preferences: Open User Settings (JSON)来执行此操作。

{
  "mcp": {
    "inputs": [
      {
        "type": "promptString",
        "id": "apify_token",
        "description": "Apify API Token",
        "password": true
      }
    ],
    "servers": {
      "actors-mcp-server": {
        "command": "npx",
        "args": ["-y", "@apify/actors-mcp-server"],
        "env": {
          "APIFY_TOKEN": "${input:apify_token}"
        }
      }
    }
  }
}

或者,您可以将其添加到工作区中名为.vscode/mcp.json的文件中 - 只需省略顶层mcp {}键即可。这样您就可以与其他人共享配置。

如果要指定加载哪些 Actors,可以添加--actors参数:

{
  "servers": {
    "actors-mcp-server": {
      "command": "npx",
      "args": [
        "-y", "@apify/actors-mcp-server",
        "--actors", "lukaskrivka/google-maps-with-contact-details,apify/instagram-scraper"
      ],
      "env": {
        "APIFY_TOKEN": "${input:apify_token}"
      }
    }
  }
}

使用 @modelcontextprotocol/inspector 调试 NPM 包 @apify/actors-mcp-server

要调试服务器,请使用MCP Inspector工具:

export APIFY_TOKEN=your-apify-token
npx @modelcontextprotocol/inspector npx -y @apify/actors-mcp-server

通过 Smithery 安装

要通过Smithery自动为 Claude Desktop 安装 Apify Actors MCP 服务器:

npx -y @smithery/cli install @apify/actors-mcp-server --client claude

Stdio 客户端

创建一个环境文件.env ,内容如下:

APIFY_TOKEN=your-apify-token

在examples目录中,您可以找到一个示例客户端,通过标准输入/输出(stdio)与服务器交互:

  • clientStdio.ts此客户端脚本使用两个指定的 Actor 启动 MCP 服务器。然后,它会通过查询调用apify/rag-web-browser工具并打印结果。它演示了如何连接到 MCP 服务器、列出可用工具以及如何使用 stdio 传输调用特定工具。

    node dist/examples/clientStdio.js

👷🏼 开发

先决条件

  • Node.js (v18 或更高版本)

  • Python 3.9 或更高版本

创建一个环境文件.env ,内容如下:

APIFY_TOKEN=your-apify-token

构建 actor-mcp-server 包:

npm run build

本地客户端(SSE)

要使用 SSE 传输测试服务器,可以使用脚本examples/clientSse.ts :目前,Node.js 客户端不支持使用自定义标头建立与远程服务器的连接。您需要在脚本中将 URL 更改为本地服务器 URL。

node dist/examples/clientSse.js

调试

由于 MCP 服务器通过标准输入/输出 (stdio) 运行,因此调试可能颇具挑战性。为了获得最佳调试体验,请使用MCP 检查器。

您可以使用以下命令通过npm启动 MCP Inspector:

export APIFY_TOKEN=your-apify-token
npx @modelcontextprotocol/inspector node ./dist/stdio.js

启动后,检查器将显示一个 URL,您可以在浏览器中访问该 URL 以开始调试。

ⓘ 限制和反馈

Actor 输入模式经过处理,兼容大多数 MCP 客户端,同时遵循JSON Schema标准。处理过程包括:

  • 描述被截断为 500 个字符(如MAX_DESCRIPTION_LENGTH中所定义)。

  • 所有元素的枚举字段被截断为最大总长度 200 个字符(如ACTOR_ENUM_MAX_LENGTH中所定义)。

  • 必填字段在其描述中明确标有“REQUIRED”前缀,以便与可能无法正确处理 JSON 模式的框架兼容。

  • 为代理配置和请求列表源等特殊情况构建嵌套属性,以确保正确的输入结构。

  • 当未在架构中明确定义时,将推断数组项类型,使用优先级顺序:项中的显式类型 > 预填充类型 > 默认值类型 > 编辑器类型。

  • 将枚举值和示例添加到属性描述中,以确保即使客户端不完全支持 JSON 模式也能看到。

每个 Actor 的内存限制为 4GB。免费用户限制为 8GB,运行Actors-MCP-Server需要分配 128MB 内存。

如果您需要其他功能或有任何反馈,请在 Apify 控制台中提交问题让我们知道。

🚀 路线图(2025 年 3 月)

  • 添加 Apify 的数据集和键值存储作为资源。

  • 添加 Actor 日志和 Actor 运行等工具以供调试。

🐛 故障排除

  • 通过运行node -v确保已安装node

  • 确保已设置APIFY_TOKEN环境变量

  • 通过设置@apify/actors-mcp-server@latest始终使用最新版本的 MCP 服务器

📚 了解更多

Available Tools

13 tools
abort-actor-runAbort Actor runA
DestructiveIdempotent
Inspect

Abort an Actor run that is currently starting or running. For runs with status SUCCEEDED, FAILED, ABORTING, ABORTED, or TIMED-OUT, this call has no effect. The results will include the updated run details after the abort request.

USAGE:

  • Use when you need to stop a run that is taking too long or misconfigured.

USAGE EXAMPLES:

  • user_input: Abort run y2h7sK3Wc

  • user_input: Gracefully abort run y2h7sK3Wc

Inputs: runId (required), gracefully. Returns: tip, runId, stats, status, +10 more

call-actorCall ActorA
Destructive
Inspect

Call any Actor from the Apify Store.

WORKFLOW:

  1. Use fetch-actor-details to get the Actor's input schema

  2. Call this tool with the actor name and proper input based on the schema

If the actor name is not in "username/name" format, use search-actors to resolve the correct Actor first.

An MCP server Actor with no input schema is not run here; call-actor-mcp-tool calls its MCP tools.

For Server Actors (Actors with an HTTP API):

  • This tool starts a batch run and stores the results in a dataset; use it when you need batch results

  • For one or a few results, send-server-actor-request returns them in the HTTP response, without a batch run

  • fetch-actor-details with output={ serverInterface: true } lists its HTTP operations

IMPORTANT:

  • Waits up to waitSecs (default 30s) for completion; returns run status and storage IDs, and with waitSecs > 0 also reports dataset field metadata

  • Use get-dataset-items with the datasetId to fetch results; non-terminal runs include a nextStep with polling instructions

There are two ways to use an Actor:

  1. Its dedicated tool, when the session has one: a Server Actor's tool sends one HTTP request to its API, an MCP server Actor's tools call its MCP server, and any other Actor's tool starts a run.

  2. The generic call-actor tool: starts a run of any Actor and stores the results in a dataset.

USAGE:

  • Prefer the dedicated tool when the session has one for the Actor

  • call-actor runs an Actor that has no dedicated tool, and batch work whose results belong in a dataset.

  • Use waitSecs (0–45) to control how long to wait. Default 30s returns results for fast actors. Use waitSecs: 0 to start and return immediately for long-running actors.

EXAMPLES:

  • user_input: Get instagram posts using apify/instagram-scraper

Inputs: actor (required), input (required), waitSecs, callOptions. Returns: tip, runId, stats, status, +10 more

call-actor-mcp-toolCall Actor MCP toolA
Destructive
Inspect

Call one tool on an Actor's MCP server and return its result. fetch-actor-mcp-tool-list lists the Actor's MCP tools and their input schemas. It connects with your Apify API token. Connecting can start a Server Actor run in your account, kept until its idle timeout; a multi-tenant Server Actor charges per event. What the tool itself does depends on the Actor.

Example call: {"actor": "username/name", "tool": "search", "arguments": {"query": "apify"}}

USAGE:

  • Use when the user asks to call a tool on an Actor's MCP server.

USAGE EXAMPLES:

  • user_input: Use the search tool of username/name to look up apify

Inputs: tool (required), actor (required), arguments.

fetch-actor-detailsFetch Actor detailsA
Read-onlyIdempotent
Inspect

Get detailed information about an Actor by its ID or full name (format: "username/name", e.g., "apify/rag-web-browser").

Requires the exact ID or full name — do not construct a plausible-looking name and call this tool with it. If you only have a description, a partial name, or a name you have not seen in this conversation, find it with search-actors first.

Use 'output' to control returned information:

  • Default: all fields except outputSchema and serverInterface. Server Actors always get a one-line server interface summary unless serverInterface is true or a selector.

  • Selective: set only the fields you need to true, e.g. output: { inputSchema: true }

  • Common patterns: inputSchema only, description + readme, serverInterface: true (every operation in full when they fit, an index of them otherwise) or "GET /search" (one operation in full) for Server Actors

The 'readme' field returns the summary when available, full README otherwise. Use when querying Actor details, documentation, or input requirements. It does not connect to an MCP server Actor; fetch-actor-mcp-tool-list lists its MCP tools.

EXAMPLES:

  • What does apify/web-fetch do?

  • What is the input schema for apify/web-scraper?

  • What HTTP operations does this Server Actor expose?

Inputs: actor (required), output. Returns: readme, nextStep, actorInfo, inputSchema, +2 more

fetch-actor-mcp-tool-listFetch Actor MCP tool listAInspect

List an Actor's MCP tools with each tool's name, description, input schema, annotations, and metadata, as the server reports them. call-actor-mcp-tool calls one of them by name. Each call asks the server again; nothing is cached. It connects with your Apify API token. Connecting can start a Server Actor run in your account, kept until its idle timeout; a multi-tenant Server Actor charges per event.

USAGE:

  • Use when the user asks which MCP tools an Actor provides, or before calling one of them.

USAGE EXAMPLES:

  • user_input: What tools does username/name provide?

Inputs: actor (required). Returns: tools

fetch-apify-docsFetch Apify docsA
Read-onlyIdempotent
Inspect

Fetch the full content of an Apify or Crawlee documentation page by its URL. Use this after finding a relevant page with the search-apify-docs tool.

USAGE:

  • Use when you need the complete content of a specific docs page for detailed answers.

USAGE EXAMPLES:

Inputs: url (required). Returns: url, content

get-actor-runGet Actor runA
Read-onlyIdempotent
Inspect

Get detailed information about a specific Actor run.

Returns run result: status, storages (datasets/keyValueStores alias map), stats, summary, nextStep.

  • summary describes the past (e.g. "SUCCEEDED in 22s. 47 items; 3 fields available.").

  • nextStep prescribes one primary follow-up action with identifiers interpolated (e.g. "Use get-dataset-items with datasetId=...").

  • waitSecs (0–45, default 30) waits up to that many seconds for terminal status before returning.

USAGE:

  • Use to check the status of a run started by any Actor-running tool.

  • Pass waitSecs > 0 to block until terminal (or until the cap elapses).

USAGE EXAMPLES:

  • user_input: Show details of run y2h7sK3Wc

  • user_input: Wait for run y2h7sK3Wc to finish

Inputs: runId (required), waitSecs. Returns: tip, runId, stats, status, +10 more

get-dataset-itemsGet dataset itemsA
Read-onlyIdempotent
Inspect

Get items (rows) from a dataset — the output/results produced by an Actor run. Returns the rows themselves, not dataset metadata, counts, or a schema. When the user provides a datasetId and asks to retrieve results, output, data, or rows, call this tool directly. Default limit is 20. Use clean=true to skip empty items and hidden fields.

USAGE:

  • Use when you need to read data from a dataset (all items or only selected fields).

USAGE EXAMPLES:

  • user_input: Retrieve results from dataset abc123

  • user_input: Get only metadata.url and title from dataset username~my-dataset

Inputs: datasetId (required), desc, omit, clean, +4 more. Returns: items, limit, offset, summary, +5 more

get-key-value-store-recordGet key-value store recordA
Read-onlyIdempotent
Inspect

Get the value stored under a specific key in a key-value store — a single record, not a listing of all keys. Requires the exact key name. The response preserves the original Content-Encoding; most clients handle decompression automatically. A record over 262144 bytes is returned as a link to download it, not as its value.

USAGE:

  • Use when you need to retrieve a specific record (JSON, text, or binary) from a store.

USAGE EXAMPLES:

  • user_input: Get record INPUT from store abc123

  • user_input: Get record data.json from store username~my-store

Inputs: recordKey (required), keyValueStoreId (required). Returns: key, value, summary, contentType, +1 more

report-problemReport a problemAInspect

Report a problem with Apify's MCP tools or Actors to the Apify team.

Reports can describe missing tools or Actors, errors, timeouts, confusing results, steps that took several attempts, or tasks that the available tools could not complete. Reports help fix the problem for other users.

Put what you were doing and what went wrong in "message". Do NOT include personal data, credentials, secrets, or verbatim private conversation content — describe the issue in your own words.

Inputs: message (required), actorId, actorRunId, relatedTools. Returns: reported

search-actorsSearch ActorsA
Read-onlyIdempotent
Inspect

Search the Apify Store to FIND and DISCOVER what scraping tools/Actors exist for specific platforms or use cases. This tool provides INFORMATION about available Actors - it does NOT retrieve actual data or run any scraping tasks.

Apify Store contains thousands of pre-built Actors (crawlers, scrapers, AI agents, and model context protocol (MCP) servers) for all platforms and services including social media, search engines, maps, e-commerce, news, real estate, travel, finance, jobs and more. Apify Store lists Actors published by Apify and by independent developers.

Use this tool when user wants to:

  • Find what scraping tools exist for a platform (e.g., "What tools can scrape Instagram?")

  • Discover available Actors for a use case (e.g., "Find an Actor for Amazon products")

  • Browse existing solutions (e.g., "Show me scrapers for news sites")

  • Learn about MCP servers or AI agents available in the Store

This tool discovers Actors; it does not retrieve their output. IMPORTANT: When the user wants detailed data from a specific platform or is looking for scraping tools, prefer searching the Store first — a relevant Actor often already exists. When the user names a specific Actor, look up that Actor directly instead of searching.

Usage:

  • Use 1-3 simple terms from the platform and/or data type the user named (e.g. "Instagram posts", "Twitter", "Amazon products"). With neither named, empty keywords return the Store's default ordering. A platform-only query is an option if more specific results do not fit.

Important limitations: This tool does not return full Actor documentation or detailed usage instructions - only summary information. Each result lists the Actor's input fields with their types (e.g. url: string, maxResults?: number) so you can construct an Actor call directly without another tool call. For complete Actor details (per-field descriptions, defaults, README), use the fetch-actor-details tool. The search is limited to publicly available Actors and excludes rental and restricted Actors.

Returns list of Actor cards with the following info:

  • Title: Markdown header linked to the Store page, followed by the full Actor name in code format

  • URL: Direct Store link

  • Description: Actor description or fallback

  • Pricing: Details with pricing link

  • Stats: Total and monthly users

  • Rating: Out of 5 (if available)

  • Developed by: Username linked to profile, marked (Apify) or (community)

  • Categories: Formatted or "Uncategorized"

  • Last modified: Date (if available)

  • Input fields: Inline list of input field names and types (e.g. url: string, maxResults?: number); ? marks optional fields, ... (+N more) marks a truncated list

Inputs: limit, offset, useCase, keywords. Returns: count, query, actors, userTier, +1 more

search-apify-docsSearch Apify docsA
Read-onlyIdempotent
Inspect

Search Apify and Crawlee documentation using full-text search. Do not also search the Apify Store unless the user asks to find Actors.

You must explicitly select which documentation source to search using the docSource parameter:

• docSource="apify" - Apify: Apify Platform documentation including: Platform features, SDKs (JS, Python), CLI, REST API, Academy (web scraping fundamentals), Actor development and deployment

• docSource="crawlee-js" - Crawlee (JavaScript): Crawlee is a web scraping library for JavaScript. It handles blocking, crawling, proxies, and browsers for you.

• docSource="crawlee-py" - Crawlee (Python): Crawlee is a web scraping library for Python. It handles blocking, crawling, proxies, and browsers for you.

The results will include the URL of the documentation page (which may include an anchor), and a limited piece of content that matches the search query.

Fetch the full content of the document using the fetch-apify-docs tool by providing the URL.

When results contain both platform documentation (docs.apify.com/platform) and Academy content (docs.apify.com/academy) on the same topic, prefer the platform documentation.

Inputs: query (required), limit, offset, docSource. Returns: results, instructions

send-server-actor-requestSend Server Actor requestA
Destructive
Inspect

Send one HTTP request to an Actor Server: the HTTP API of a Server Actor, documented in its API schema (OpenAPI) and README. fetch-actor-details with output.serverInterface returns the documented operations with their parameters. Give the method, the path, and optionally query parameters and a JSON body; the Actor Server's host and your Apify API token are added for you. Only an Actor with an Actor Server and an API schema takes requests here; the Actor accepts or rejects the method and path. A request to a single-tenant Actor Server can start a run of the Actor in your account, kept until its idle timeout; a multi-tenant Actor Server charges per event. The request is sent once, with no retries and without following redirects. A JSON, text, or XML response body is returned as text, cut at 262144 bytes; other content types are not returned.

Example call: {"actor": "apify/rag-web-browser", "method": "GET", "path": "/search", "query": {"query": "AI news", "maxResults": 3}}

USAGE:

  • Use when the Actor that fits the user's request has an Actor Server, and its API schema or README names the endpoint for the task.

  • Fits requests for one or a few results returned in the HTTP response. Larger batches depend on the endpoint and the response size and time limits.

USAGE EXAMPLES:

  • user_input: Search the web for the latest AI news and read the top pages with apify/rag-web-browser

  • user_input: Read https://docs.apify.com as Markdown with apify/rag-web-browser

Inputs: path (required), actor (required), method (required), body, +1 more. Returns: body, path, bytes, method, +13 more

Tool Schema Changelog

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

Loading tool schema changelog…

    TDQS

    A4.3/5.0

    Scored across 13 tools

    Disambiguation4/5

    Most tools target clearly distinct actions (search, fetch details, run, abort, read dataset, docs search/fetch). The only murky cluster is call-actor vs call-actor-mcp-tool vs send-server-actor-request, which all 'invoke' something, but the descriptions clearly differentiate generic runs, MCP tool calls, and HTTP endpoint requests.

    Naming Consistency5/5

    All 13 tools use consistent kebab-case verb-first naming: fetch-*, search-*, get-*, call-*, abort-*, send-*, report-*. No mixing of conventions or casing styles.

    Tool Count5/5

    13 tools sits squarely in the well-scoped range for an Actor hub that must cover discovery, execution, storage retrieval, docs, and reporting. Each tool maps to a distinct step in the workflow.

    Completeness4/5

    The surface covers the full Actor lifecycle (discover, inspect, run, poll, abort) plus dataset and key-value access, docs, MCP tool invocation, server HTTP requests, and problem reporting. Minor gaps remain: no listing of a user's own runs/actors or dataset/key-value-store enumeration, though agents can work around these with known IDs.

    Maintenance

    ActivityNo data
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Exposes over 19,000 Apify Actors as MCP tools for web scraping, data extraction, and OSINT automation. It enables AI agents to dynamically discover and execute scrapers to collect structured data and crawl web content.
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      35 public-data MCP tools with optional focused profiles for business leads, market intelligence, government records, and research/health. Includes Maps, jobs, SEC filings, grants, sanctions and more. Runs use your own Apify account and are billed there. Structured results and deterministic errors.
      43
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables AI agents to access fourteen web-data tools for scraping, SEO and site auditing, extraction, validation, enrichment, and screenshots through Apify actors. It runs those tools on the user's own Apify account via an API token.
      MIT