Tasks
Tasks represent units of work that your system executes—whether on-demand, scheduled, or triggered by external events. They serve as the backbone for automating processes and tracking their performance over time.
Key Concepts:
-
Core Attributes:
- Identification: Each task has a unique ID and a name that distinguishes it.
- Description & Instructions: These fields explain what the task does and provide detailed guidance for its execution.
- Model & Response Type: Specify the AI model used and the expected response format (e.g., JSON), ensuring consistent output.
- Temperature: A value that controls the randomness in the task’s response.
-
Execution and Monitoring:
- Metrics: Tasks log performance data such as total calls, completions, failures, and average execution time (displayed in reports).
- Scheduling: Tasks can be set to run automatically using scheduling properties like interval minutes, cron expressions, and defined start/end dates.
- Notifications: Options for email and webhook notifications keep you informed when tasks execute or encounter issues.
-
Interconnections:
- Functions & Files: Tasks may be linked to various functions (which perform specialized operations) and files (assets or data sources) to support their execution.
- Completion References: These reference additional tasks or subtasks that contribute to the overall completion of a process.
User Benefits:
By organizing processes into tasks, the system enables automation and continuous monitoring. This structure provides clarity on performance metrics, helps schedule recurring work, and facilitates integration with other system components, ensuring that complex workflows run reliably.