MewCP Apify MCP

Run web scrapers, automate data extraction, and manage Actor pipelines through AI.

A Model Context Protocol (MCP) server that exposes Apify's API for running Actors, managing runs, and retrieving scraped datasets.

The Apify MCP Server provides end-to-end control over Apify's automation platform through AI:

- Discover and run Actors from your Apify account with custom inputs and resource limits
- Monitor run status and retrieve scraped output from datasets
- Browse and manage pre-configured Actor tasks

- Triggering web scraping pipelines from conversational AI interfaces
- Polling run results and surfacing structured dataset output without leaving the chat
- Managing Actor task libraries and monitoring run history through natural language

Returns a status object confirming the server is running and reachable.

{ "status": "ok", "server": "CL Apify MCP Server" }

Returns a paginated list of Actors in your Apify account, including ID, name, and username.

- my_only (boolean, optional) — Only return Actors owned by you (default: true) - limit (integer, optional) — Maximum number of Actors to return, 1–1000 (default: 100) - offset (integer, optional) — Number of Actors to skip for pagination (default: 0)

{ "success": true, "total": 12, "count": 12, "actors": [ { "id": "abc123", "name": "web-scraper", "username": "myuser" } ] }

Starts an Actor run with the provided input and optional resource constraints. Returns the run ID and default dataset ID for polling results.

- actor_id (string, required) — Actor ID to run (e.g. 'username~actor-name' or Actor ID) - input_data (string, optional) — JSON string of input data for the Actor (default: '{}') - timeout_secs (integer, optional) — Run timeout in seconds - memory_mbytes (integer, optional) — Memory limit in MB (min 128) - build (string, optional) — Build tag or number (default: 'latest')

{ "success": true, "run_id": "run_XXXXXXXXXX", "status": "RUNNING", "started_at": "2024-01-01T00:00:00.000Z", "default_dataset_id": "dataset_XXXXXXXXXX", "default_key_value_store_id": "kvs_XXXXXXXXXX" }

Retrieves the status and metadata of a specific Actor run. Optionally waits up to 60 seconds for the run to finish.

- run_id (string, required) — Actor run ID - wait_for_finish (integer, optional) — Seconds to wait for run completion, 0–60 (default: 0)

{ "success": true, "run_id": "run_XXXXXXXXXX", "actor_id": "abc123", "status": "SUCCEEDED", "started_at": "2024-01-01T00:00:00.000Z", "finished_at": "2024-01-01T00:01:30.000Z", "default_dataset_id": "dataset_XXXXXXXXXX", "usage_total_usd": 0.012 }

Returns a paginated list of Actor runs from your account, sorted newest first. Optionally filter by run status.

- status (string, optional) — Filter by status: 'SUCCEEDED', 'FAILED', 'RUNNING', 'ABORTED', etc. - limit (integer, optional) — Maximum number of runs to return, 1–1000 (default: 100) - offset (integer, optional) — Number of runs to skip for pagination (default: 0)

{ "success": true, "total": 42, "count": 10, "runs": [ { "id": "run_XXXXXXXXXX", "actor_id": "abc123", "status": "SUCCEEDED", "started_at": "2024-01-01T00:00:00.000Z", "finished_at": "2024-01-01T00:01:30.000Z" } ] }

Fetches scraped items from an Actor run's default dataset. This is the primary way to read Actor output after a run completes.

- dataset_id (string, required) — Dataset ID (returned in the Actor run response) - limit (integer, optional) — Maximum number of items to return, 1–10000 (default: 100) - offset (integer, optional) — Number of items to skip for pagination (default: 0) - clean (boolean, optional) — Remove hidden fields starting with '#' (default: true)

{ "success": true, "count": 25, "items": [ { "url": "https://example.com", "title": "Example Page", "price": 29.99 } ] }

Returns a paginated list of Actor tasks in your account. Tasks are pre-configured Actor runs with saved inputs.

- limit (integer, optional) — Maximum number of tasks to return, 1–1000 (default: 100) - offset (integer, optional) — Number of tasks to skip for pagination (default: 0)

{ "success": true, "total": 5, "count": 5, "tasks": [ { "id": "task_XXXXXXXXXX", "name": "my-scraper-task", "actor_id": "abc123", "username": "myuser", "created_at": "2024-01-01T00:00:00.000Z" } ] }

- limit— Maximum number of records to return per request (max varies by endpoint)
- offset— Number of records to skip; use withlimitfor pagination
- desc— Sort order; list endpoints return results newest-first by default

{username}~{actor-name} or {actorId} Example: apify~web-scraper or BwFbCCmwYxNqHr7TB

- READY— Queued and waiting to start
- RUNNING— Currently executing
- SUCCEEDED— Completed successfully
- FAILED— Terminated with an error
- ABORTING— Abort in progress
- ABORTED— Stopped by user or timeout
- TIMED-OUT— Exceeded the timeout limit
- Go to theApify Console
- Click your profile avatar →Settings→Integrations
- UnderAPI tokens, click+ Add new token
- Give the token a name and clickCreate— copy the token value immediately, it is only shown once

- Cause:Upstream Apify API returned an error
- Solution:
- Check Apify service status atApify Status Page
- Verify your API token has the required permissions for the operation
- Review the error message for specific details (e.g. Actor not found, insufficient compute units)

- Apify API Documentation— Official API reference
- Apify Console— Manage Actors, runs, and datasets
- FastMCP Docs— FastMCP specification
- FastMCP Credentials— FastMCP Credentials package for credential handling

Build and deploy full-stack Next.js apps with 98 tools for React, AWS, and MongoDB