Skip to content
Nappai Documentation

JsonAgent

This component lets Nappai understand and use information stored in structured data files like JSON or YAML. Think of it as a translator that helps Nappai understand your data and perform actions based on it. It uses a powerful language model to interpret the instructions in your data file.

Relationship with Language Models

This component works with a language model (like those from OpenAI or Google). The language model provides the intelligence to understand the instructions and data within your JSON or YAML file. The JsonAgent uses this model to process the information and perform the actions specified in your file.

Inputs

  • Language Model: This is the language model Nappai will use to understand your data. You need to select a language model from the available options. This is required.
  • File Path: This is the location of your JSON or YAML file. Nappai needs this file to know what actions to perform. This is required. The file must be in either JSON, YAML, or YML format.

Outputs

The JsonAgent doesn’t have named outputs in the traditional sense. Instead, it creates an internal “agent” that can perform the tasks defined in your JSON or YAML file. The results of these tasks will then be available through other components connected to the JsonAgent in your workflow. For example, if your JSON file instructs the agent to send an email, the email sending will be handled by a separate email component.

Usage Example

Imagine you have a JSON file that lists customer orders. You can use the JsonAgent to process this file. By connecting the JsonAgent to other Nappai components, you could automatically:

  1. Identify new orders.
  2. Send confirmation emails to customers.
  3. Update an inventory database.

All this happens automatically based on the instructions in your JSON file and the power of the chosen language model.

Templates

[This section would list any pre-built templates in Nappai that utilize the JsonAgent component. Since none are provided, this section remains empty.]

Many components can work with the JsonAgent to complete tasks. Here are a few examples:

  • Email Sender: Sends emails based on information processed by the JsonAgent.
  • Database Updater: Updates databases with information extracted by the JsonAgent.
  • File Writer: Creates or updates files based on the JsonAgent’s processing.
  • Many other components: The JsonAgent can be used with almost any component that can receive data as input.

Tips and Best Practices

  • Clear File Structure: Make sure your JSON or YAML file is well-organized and easy to understand. The clarity of your file directly impacts the accuracy of the automation.
  • Specific Instructions: Provide clear and concise instructions in your JSON or YAML file. Ambiguous instructions can lead to errors.
  • Error Handling: Consider how to handle potential errors (e.g., missing data in the file). You might want to use other Nappai components to manage errors gracefully.

Security Considerations

  • Data Sensitivity: If your JSON or YAML file contains sensitive information (e.g., customer data, financial information), ensure that your Nappai workflow and the chosen language model are configured securely to protect this data. Review Nappai’s security documentation for best practices.
  • Access Control: Restrict access to your JSON files and the Nappai workflow to authorized personnel only.