Drafting/Debugging an Agent
Save Draft
The Save Draft option in the Sequentum Cloud status bar allows users to save their created agent as a draft, similar to email functionality. After choosing this option, the agent will be stored in the Drafts folder within the Control Center.
When the user clicks on the Save Draft option available in the Sequentum Cloud Editor, a text box will appear on the screen indicating three options Config name, Save and Cancel. Below are the screenshots and descriptions for the same.
Config name: Users can assign any desired name for the agent in this field, which will also appear in the Control Center under the Drafts folder. By default, the value is set to “New Agent”.
Save: This option will save the Draft and the same will be visible in the Drafts folder in Sequentum Cloud Control Center.
Cancel: This option cancels the Draft Saving process, closes the Save Draft dialog box, and returns to design mode.
After entering a name in the Save Draft dialog box and clicking the Save button, as shown in the screenshot below, the agent will be saved as a draft and will appear in the Drafts folder within the Sequentum Cloud Control Center.
Once the draft is saved, click on the Control Center option in the top right corner of the Sequentum Editor window. A pop-up message will appear, presenting four options:
Publish the agent
Keep Draft
Discard and Exit
Cancel to return to design mode
Selecting the Keep Draft option will save the draft in the Drafts folder under the Control Center, as illustrated in the screenshot below.
Play
The Play button, located in the status bar of the Sequentum Cloud Editor, is a crucial feature for users. Clicking this button initiates debugging from the start of the agent. The agent debugger is an essential tool for identifying and resolving issues, and running the agent through the debugger is typically the first step after building it.
Once the user clicks the Play button and debugging begins, two new options will appear in the status bar:
Stop: This option allows the user to end the debugging process at any node whenever the user wants to stop in between.
Play Speed: This scrollbar lets users control the debugging speed. By default, the speed is set to normal (in the middle of the scrollbar), but users can slow down the debugging speed to observe the agent's actions more closely during execution or increase the speed for faster processing.
Reload
The Reload feature in the Sequentum Cloud Editor is designed to reload the agent or configuration settings by clicking the reload button in the configuration menu. This functionality is particularly useful when a reload is required due to updates or changes made to the agent or its configuration. By reloading, the system refreshes the current state of the agent, ensuring that the latest settings and configurations are applied.
However, it is crucial to note that before using the Reload feature, users must save any changes made to the agent by using the Save Draft option. If changes are not saved before reloading, the system will discard unsaved modifications and all the progress will be lost. Therefore, saving your work before reloading is essential to prevent unintentional data loss or the need to redo the changes.
The Reload feature ensures that the agent is operating with the most current configurations, making it an important step during development or after significant changes have been made. It helps in resetting the agent to reflect new updates, configurations, or debugging steps. This function is particularly helpful during troubleshooting, testing or when switching between different configurations, as it enables developers to reload the environment quickly and efficiently.
In summary, the Reload button provides a convenient way to refresh the agent, but it must be used with caution. Always ensure that changes are saved before reloading to avoid losing any important updates.
Play Log
The Play Log feature in the Sequentum Cloud Editor is a powerful tool used to check the debug time logs during the agent's debugging process. This feature allows developers to monitor and review the detailed logs generated while the agent is being executed in debug mode. By providing a comprehensive view of each step and the details about the command the agent executes, the Play Log helps to identify any potential issues or anomalies in real-time.
With the Play Log, users can track the exact sequence of actions the agent performs, along with the time taken for each command. This information is invaluable for pinpointing bottlenecks, errors or performance issues that may arise during the debugging process. It essentially provides a full breakdown of the agent's operations, offering visibility into the success or failure of individual tasks.
Moreover, the Play Log allows developers to understand the flow of data and commands within the agent, helping them to trace any problem areas and make necessary adjustments. Whether it's a misconfiguration, incorrect logic or an external integration issue, the logs serve as a detailed guide for troubleshooting and optimizing the agent's performance.
In summary, the Play Log feature is essential for efficient debugging, offering developers a clear, detailed overview of how the agent runs step by step. It ensures that any errors or inefficiencies are quickly identified and resolved, leading to more stable and reliable agent performance.
Play Data
When we start the agent in play mode for debugging by clicking the play button, the agent runs through its defined tasks and interactions and the debug process begins. Once the agent completes its execution, and we want to review the crawled or extracted data, we can access this information in the "Play Data" tab. The Play Data tab is located right after the Play Log tab, making it easy to navigate between the logs and the actual data that the agent has gathered during the debugging session.
The Play Data tab provides a clear, organized view of the data collected during the agent's execution. This feature is especially useful for verifying whether the agent has successfully retrieved the expected data from the target sources. It displays all the collected data in a structured format, allowing developers to inspect the accuracy, completeness and formatting of the extracted information.
By reviewing the data in the Play Data tab, developers can ensure that the agent is working correctly and that the data extraction process is functioning as intended. If any discrepancies or missing data are identified, developers can quickly make adjustments to the agent’s configuration or logic and rerun the debugging process. This iterative approach ensures that the agent performs optimally before being deployed in a live environment.
In summary, the Play Data tab is an essential tool for reviewing and validating the crawled data during the debugging process. It allows developers to easily cross-check the results with the Play Log and make informed decisions about any further improvements or changes needed for the agent.
Schema
Play also provides a Schema View within the play tabs. This feature allows developers to easily visualize the schema of the extracted columns. By displaying the structure of the data in a clear and organized manner, it helps developers quickly understand the relationships between columns, identify data types and spot any discrepancies. This enhanced visibility of the schema ensures smoother debugging and testing processes, enabling more efficient data manipulation and validation during development.
Play Until Selected Command
The "Play Until Selected Command" feature is a valuable debugging tool in the Sequentum Cloud Editor, designed to allow users to run the agent up to a specific point or command in its execution. This feature is particularly useful when developers want to focus on debugging a specific section of the agent without having to run the entire process.
For instance, in the screenshot provided, the login command has been selected. When using the "Play Until Selected Command" feature, enabling developers to run the agent up to a desired point in its workflow. At this point, the agent will pause and exit, allowing the user to review the state of the agent just before executing the login action.
This functionality is extremely helpful in scenarios where users need to isolate and troubleshoot particular sections of the workflow. It allows developers to examine the agent's behavior leading up to critical steps, such as data input, authentication or API calls without having to step through unnecessary parts of the process. This targeted debugging approach helps save time and resources by focusing on the exact point where issues may arise.
Additionally, after running the agent up to the selected command, users can review the Play Log and Play Data to see the results up to that point providing further insight into how the agent is performing. By leveraging this feature, developers can efficiently troubleshoot, test and refine specific parts of the agent's execution, improving the overall accuracy and functionality of the agent.
In summary, the "Play Until Selected Command" feature allows for more precise and efficient debugging by enabling developers to run the agent up to a desired point in its workflow. It simplifies the process of identifying and resolving issues in complex agents by narrowing down the focus to a specific command or action.