Actors are serverless cloud programs that run on Apify's infrastructure. They perform web scraping, data extraction, and automation tasks. The agent can run any Actor by providing its ID (like 'apify/web-scraper') and input parameters. The agent retrieves results from the Actor's default dataset or key-value store output.

Yes. The Apify Store has 10,000+ ready-made Actors for popular sites like Google Maps, Amazon, LinkedIn, Instagram, and Twitter. The agent can run any public Actor by specifying its username and name (e.g., 'apify/google-search-scraper') along with the required input fields. No coding needed - just provide search terms, URLs, or other inputs.

For long-running jobs, the agent runs Actors asynchronously and polls for completion. It can set waitForFinish up to 60 seconds per request, then retrieve the run status and continue waiting. For very long jobs, the agent stores the run ID and checks back later. It can also set webhooks to receive notifications when runs complete.

Yes. The agent creates schedules using standard cron syntax (like '0 8 * * *' for 8 AM daily) or shortcuts like @daily, @weekly, and @monthly. It configures the target Actor or task, input parameters, memory allocation, and timezone. Schedules can be set to exclusive mode to prevent overlapping runs if a previous execution is still running.

The agent fetches dataset items with support for JSON, CSV, XML, Excel, and RSS formats. It can filter specific fields, skip empty items, apply sorting, paginate through large result sets (up to 250,000 items), and flatten nested JSON. For example, after running a product scraper, the agent can retrieve only price and title fields sorted by price.

Yes. The agent creates named request queues, adds URLs individually or in batches of 25, and retrieves pending requests for processing. It supports request locking for distributed crawlers where multiple workers process the same queue, handles retry logic, and marks requests as handled after successful processing.

The agent can resurrect failed or timed-out runs to continue from the last checkpoint, preserving already scraped data. It retrieves execution logs to diagnose blocking issues. For runs that need to be stopped, the agent can abort gracefully, giving the Actor 30 seconds to save state before force-stopping. This enables resuming later.

The integration uses Apify API tokens passed in the Authorization header as Bearer tokens. You generate a token in the Apify Console under Integrations and add it to Tars. The token grants access to your Actors, datasets, key-value stores, and schedules. API rate limits are 250,000 requests per minute globally and 60 requests per second per resource.