Using API Agent

API Agent helps you manage the API lifecycle by searching metadata, generating OpenAPI documents, creating and deploying Python code, and validating APIs before publishing.

You can use API Agent for end-to-end API publishing and validating. You can learn what to expect when using the API Agent plug-in and review a sample end-to-end flow that demonstrates how the API Agent works and the capabilities it provides.

Assign roles to enable API Agent for selected developers. After configuration, developers can install the Visual Studio Code plug-in and integrate it into their workspace. For more information, see the .

The following sections describe how to use API Agent to do key actions, including searching APIs and tables for metadata, generating OpenAPI documents from existing sources, creating Python code from OpenAPI specifications, deploying applications to IBM Code Engine, and creating, publishing, and validating APIs.

Important: When you interact with the agent, it generates and runs the plan step by step. For each step, the agent displays the next tool that it intends to run and requests your approval. The interface also shows the tools that were run for previous requests in the same session. This view helps you track the complete sequence of actions.

Workaround: You can create a new session to avoid the accumulating of the previous plan steps to have a better display of the results.