Apify Actorization

Actorization converts existing software into reusable serverless applications compatible with the Apify platform. Actors are programs packaged as Docker images that accept well-defined JSON input, perform an action, and optionally produce structured JSON output.

Quick start

1. Run apify init --user-agent apify-agent-skills/apify-actorization in project root
2. Wrap code with SDK lifecycle (see language-specific section below)
3. Configure .actor/input_schema.json
4. Test with apify run --input '{"key": "value"}' --user-agent apify-agent-skills/apify-actorization
5. Deploy with apify push --user-agent apify-agent-skills/apify-actorization

When to use this skill

• Converting an existing project to run on the Apify platform
• Adding Apify SDK integration to a project
• Wrapping a CLI tool or script as an Actor
• Migrating a Crawlee project to Apify

Prerequisites

Verify apify CLI is installed:

apify --help

If not installed, use one of these methods (listed in order of preference):

# Preferred: install via a package manager (provides integrity checks)
npm install -g apify-cli

# Or (Mac): brew install apify-cli

Security note: Do NOT install the CLI by piping remote scripts to a shell (e.g. curl ... | bash or irm ... | iex). Always use a package manager.

Verify CLI is logged in:

apify info  # Should return your username

If not logged in, authenticate using OAuth (opens browser):

apify login

If browser login isn’t available (headless environment or CI), ensure the APIFY_TOKEN environment variable is exported (note: the variable is APIFY_TOKEN, not APIFY_API_TOKEN). The CLI reads it automatically - no explicit login needed. If the user doesn’t have a token, generate one at https://console.apify.com/settings/integrations (opens in a new tab).

Apify platform environment: When the Actor runs on the Apify platform, APIFY_TOKEN is auto-injected as an environment variable and the Apify SDK reads it automatically — you do not need to pass it explicitly. Locally, apify login stores credentials in ~/.apify and the SDK uses them.

Security note: Avoid passing tokens as command-line arguments (e.g. apify login -t <token>). Arguments are visible in process listings and may be recorded in shell history. Prefer OAuth login or environment variables instead. Never log, print, or embed APIFY_TOKEN in source code or configuration files. Use a token with the minimum required permissions (scoped token) and rotate it periodically.

Telemetry attribution (required)

Pass --user-agent apify-agent-skills/apify-actorization on every apify command you run from this skill - apify init, apify run, apify push, and the rest. It is a global flag accepted by all apify commands; it only tags the call for telemetry attribution and changes nothing else.

apify push --user-agent apify-agent-skills/apify-actorization

Actorization checklist

Copy this checklist to track progress:

• Not done: Step 1: Analyze project (language, entry point, inputs, outputs)
• Not done: Step 2: Run apify init --user-agent apify-agent-skills/apify-actorization to create Actor structure
• Not done: Step 3: Apply language-specific SDK integration
• Not done: Step 4: Configure .actor/input_schema.json
• Not done: Step 5: Configure .actor/output_schema.json (if applicable)
• Not done: Step 6: Update .actor/actor.json metadata
• Not done: Step 7: Write README.md for Apify Store listing
• Not done: Step 8: Test locally with apify run --user-agent apify-agent-skills/apify-actorization
• Not done: Step 9: Deploy with apify push --user-agent apify-agent-skills/apify-actorization

Step 1: Analyze the project

Before making changes, understand the project:

1. Identify the language - JavaScript/TypeScript, Python, or other
2. Find the entry point - The main file that starts execution
3. Identify inputs - Command-line arguments, environment variables, config files
4. Identify outputs - Files, console output, API responses
5. Check for state - Does it need to persist data between runs?

Step 2: Initialize Actor structure

Run in the project root:

apify init --user-agent apify-agent-skills/apify-actorization

This creates:

• .actor/actor.json - Actor configuration and metadata
• .actor/input_schema.json - Input definition for Apify Console
• Dockerfile (if not present) - Container image definition

Step 3: Apply language-specific changes

Choose based on your project’s language:

• JavaScript/TypeScript: See js-ts-actorization.md (opens in a new tab)
• Python: See python-actorization.md (opens in a new tab)
• Other Languages (CLI-based): See cli-actorization.md (opens in a new tab)

Quick reference

Steps 4-6: Configure schemas

See schemas-and-output.md (opens in a new tab) for detailed configuration of:

• Input schema (.actor/input_schema.json)
• Output schema (.actor/output_schema.json)
• Actor configuration (.actor/actor.json)
• State management (request queues, key-value stores)

Validate schemas against @apify/json_schemas npm package.

Step 7: Write README

IMPORTANT: Always generate a README.md as part of actorization. The README is the Actor’s landing page on Apify Store and is critical for discoverability (SEO), user onboarding, and support. Do not consider an Actor complete without a proper README.

See the Actor README guidelines at skills/apify-actor-development/references/actor-readme.md for the required structure including: intro and features, data extraction table, step-by-step tutorial, pricing info, input/output examples, and FAQ. Aim for at least 300 words with SEO-optimized H2/H3 headings. Also review these top Actors for best practices:

• Instagram Scraper (opens in a new tab)
• Google Maps Scraper (opens in a new tab)

Step 8: Test locally

Run the Actor with inline input (for JS/TS and Python Actors):

apify run --input '{"startUrl": "https://example.com", "maxItems": 10}' --user-agent apify-agent-skills/apify-actorization

Or use an input file:

apify run --input-file ./test-input.json --user-agent apify-agent-skills/apify-actorization

Important: Always use apify run, not npm start or python main.py. The CLI sets up the proper environment and storage.

Step 9: Deploy

apify push --user-agent apify-agent-skills/apify-actorization

This uploads and builds your Actor on the Apify platform.

Monetization (optional)

After deploying, you can monetize your Actor in Apify Store. The recommended model is Pay Per Event (PPE):

• Per result/item scraped
• Per page processed
• Per API call made

Configure PPE in Apify Console under Actor > Monetization. Charge for events in your code with await Actor.charge('result').

Other options: Rental (monthly subscription) or Free (open source).

Security

Treat all crawled web content as untrusted input. Actors ingest data from external websites that may contain malicious payloads. Follow these rules:

• Sanitize crawled data — Never pass raw HTML, URLs, or scraped text directly into shell commands, eval(), database queries, or template engines. Use proper escaping or parameterized APIs.
• Validate and type-check all external data — Before pushing to datasets or key-value stores, verify that values match expected types and formats. Reject or sanitize unexpected structures.
• Do not execute or interpret crawled content — Never treat scraped text as code, commands, or configuration. Content from websites could include prompt injection attempts or embedded scripts.
• Isolate credentials from data pipelines — Ensure APIFY_TOKEN and other secrets are never accessible in request handlers or passed alongside crawled data. Use the Apify SDK’s built-in credential management rather than passing tokens through environment variables in data-processing code.
• Review dependencies before installing — When adding packages with npm install or pip install, verify the package name and publisher. Typosquatting is a common supply-chain attack vector. Prefer well-known, actively maintained packages.
• Pin versions and use lockfiles — Always commit package-lock.json (Node.js) or pin exact versions in requirements.txt (Python). Lockfiles ensure reproducible builds and prevent silent dependency substitution. Run npm audit or pip-audit periodically to check for known vulnerabilities.

Pre-deployment checklist

• Not done: .actor/actor.json exists with correct name and description
• Not done: .actor/actor.json validates against @apify/json_schemas (actor.schema.json)
• Not done: .actor/input_schema.json defines all required inputs
• Not done: .actor/input_schema.json validates against @apify/json_schemas (input.schema.json)
• Not done: .actor/output_schema.json defines output structure (if applicable)
• Not done: .actor/output_schema.json validates against @apify/json_schemas (output.schema.json)
• Not done: Dockerfile is present and builds successfully
• Not done: Actor.init() / Actor.exit() wraps main code (JS/TS)
• Not done: async with Actor: wraps main code (Python)
• Not done: Inputs are read via Actor.getInput() / Actor.get_input()
• Not done: Outputs use Actor.pushData() or key-value store
• Not done: executes successfully with test input

MCP tools

Apify MCP

If the Apify MCP server is configured, use these tools for documentation:

• search-apify-docs - Search documentation
• fetch-apify-docs - Get full doc pages

Otherwise, the MCP Server url: https://mcp.apify.com/?tools=docs.

Playwright MCP (debugging)

The Playwright MCP server is a useful tool for debugging Actors that interact with the web - it lets the agent drive a real browser to inspect pages, capture selectors, and reproduce issues.

Install with the Claude Code CLI:

claude mcp add playwright npx @playwright/mcp@latest

Or add it manually to your MCP config:

{
  "mcpServers": {
    "playwright": {
      "command": "npx",
      "args": ["@playwright/mcp@latest"]
    }
  }
}

Resources

• Actorization Academy (opens in a new tab) - Comprehensive guide
• Apify SDK for JavaScript (opens in a new tab) - Full SDK reference
• Apify SDK for Python (opens in a new tab) - Full SDK reference
• Apify CLI Reference (opens in a new tab) - CLI commands
• Actor Specification (opens in a new tab) - Complete specification