Skip to main content
This guide explains how to use the Python SDK to upload a video for automatic task generation and check the processing status. Prerequisites: Before using the methods described here, ensure you have:
  1. Followed the Python SDK Setup guide for installation and initialization.
  2. Obtained your credentials as described in Managing API Keys.
  3. Successfully initialized the agent object as shown in the SDK Setup guide.
This feature allows you to generate a task definition by providing a video recording of the desired actions. The processing happens asynchronously.

​
Uploading Video for Task Generation

Upload a video file and associate it with an end-user to start the task generation process. 
import os
from curalabs_agent import cura # Assuming 'cura' is the client object

# Assume agent is initialized as shown in the Initialization section
# agent = cura(...)

user_id = "user-unique-id-in-your-system-789" # Replace with the end-user's ID in your system
video_file_path = "path/to/your/screen_recording.mp4" # Path to the video file on your system
optional_task_name = "Task from Recording" # Optional: Provide a name

try:
    # Note: The SDK likely handles opening and sending the file content.
    # Ensure the file path is correct and accessible.
    job_details = agent.create_task_from_video( # Updated method name
        developer_user_id=user_id, # Required: Your internal ID for the end-user
        file_path=video_file_path, # Required: Path to the video file
        task_name=optional_task_name # Optional: Name for the generated task
    )

    if job_details:
        job_id = job_details.get('job_id')
        print(f"Video upload accepted. Processing job started with ID: {job_id}")
        # Store job_id to check status later
    else:
        print("Failed to start video processing job.")

except FileNotFoundError:
    print(f"Error: Video file not found at {video_file_path}")
except Exception as e:
    print(f"An error occurred while uploading the video: {e}")
The create_task_from_video method requires the developer_user_id and the file_path to the video. It’s assumed the SDK handles reading the file content from the provided path. You can also provide an optional task_name for the generated task. On success, it returns a dictionary containing the job_id for this processing task and its initial status (usually ‘queued’).

​
Checking Processing Job Status (Polling)

Monitor the status of the asynchronous video-to-task generation job using the job_id. 
import time

job_id_to_check = "vwp_..." # Replace with the actual job_id obtained previously

if job_id_to_check:
    while True:
        try:
            status_info = agent.get_task_processing_job_status(job_id_to_check) # Updated method name

            if not status_info:
                print("Failed to get video processing job status.")
                break

            status = status_info.get('status')
            print(f"Current job status: {status}")

            if status == 'completed':
                print("Video processing completed.")
                generated_task_id = status_info.get('task_id') # Updated field name
                print(f"Generated Task ID: {generated_task_id}")
                # You can now use this task_id
                break
            elif status == 'failed':
                print("Video processing failed.")
                print(f"Error: {status_info.get('error')}")
                break
            elif status in ['queued', 'processing']:
                # Wait before checking again
                print("Job still in progress, checking again in 15 seconds...")
                time.sleep(15) # Processing might take longer
            else:
                print(f"Unknown status: {status}")
                break

        except Exception as e:
            print(f"An error occurred while checking job status: {e}")
            break
else:
    print("No Job ID provided to check status.")
The get_task_processing_job_status method requires the job_id. It returns a dictionary containing the job’s current status. If completed, it includes the task_id of the generated task definition. If failed, it includes an error message. It may also contain timestamps. Returns None on error.