Skip to content
Nappai Documentation

OpenAPI Agent

This component acts as a bridge between Nappai and various online services that use a standard way of describing their data called OpenAPI. It allows Nappai to automatically access and use information from these services without needing to write complex code.

Relationship with OpenAPI and LLMs

This component uses a special type of program called a “Language Model” (LLM) to understand and interact with APIs that follow the OpenAPI standard. The LLM helps Nappai understand what data to request and how to interpret the responses from the API. The component also uses tools specifically designed to handle the technical details of communicating with these APIs.

Inputs

  • Language Model: This is the “brain” of the component. You need to select a Language Model from the available options in Nappai. This model will be used to understand your requests and communicate with the API. This is a required input.
  • File Path: This is the location of a file that describes the API you want to use. This file is usually in JSON or YAML format and tells Nappai how to communicate with the API. This is a required input.
  • Allow Dangerous Requests: This is a safety switch. By default, it’s set to “False” (off), preventing Nappai from making requests that could potentially cause problems. You can turn it “on” (True) if you are sure the API is safe and you understand the risks. This is a required input.

Outputs

This component doesn’t directly produce visible outputs in the dashboard. Instead, it creates an internal “agent” that Nappai uses to interact with the chosen API. The results of the API interaction will be shown by other components further down your workflow in Nappai.

Usage Example

Imagine you want to use an online weather API to get the current temperature. You would:

  1. Select the OpenAPI Agent component in your Nappai workflow.
  2. Choose a Language Model.
  3. Provide the file path to the weather API’s OpenAPI description file.
  4. Leave “Allow Dangerous Requests” set to “False” unless you have specific reasons to change it.
  5. Connect the OpenAPI Agent to other Nappai components that will use the weather data (e.g., a component to display the temperature on your dashboard).

Nappai will then use the OpenAPI Agent to communicate with the weather API and retrieve the temperature information.

Templates

[This section will be populated with links to specific workflow templates once they are available.]

[This section will be populated with links to other relevant Nappai components and brief descriptions once the full system is defined. Examples include components that might use the output of the OpenAPI Agent, such as data display components or other API interaction tools.]

Tips and Best Practices

  • Make sure the OpenAPI specification file is correct and points to a working API.
  • Start with “Allow Dangerous Requests” set to “False” for security.
  • Carefully choose a Language Model appropriate for the complexity of the API interaction.

Security Considerations

Always be cautious when using the “Allow Dangerous Requests” option. Enabling this option could allow Nappai to make requests that could compromise your data or system security. Only enable this option if you fully understand the implications and trust the API you are interacting with.