AI linting support for API specifications

AI linting evaluates API specifications for quality, correctness, and consistency.

You can view AI-powered code recommendations in the AI Enhancements tab and by hovering over suggestions in the editor. These recommendations help improve your OpenAPI specifications by highlighting best practices, identifying common issues, and suggesting implementation improvements.

The AI Enhancements tab highlights areas in your API definition that can be improved. It provides suggestions on syntax and standard practice guidance to improve design consistency and overall API quality.

AI-powered recommendations include the following details:

  • Severity. Indicates whether the issue is an error or a warning.
  • Message. Specifies where the issue occurs in the code and provides a brief explanation.

Click Launch AI to open the AI Enhancements tab. Then click Generate enhancements to view suggestions based on your API definition.

You can take the following actions on each recommendation:

  • Apply fix. Automatically update the specification using the suggested improvement, if available.
  • Ignore. Dismiss a recommendation if it does not apply to your use case.
  • View explanation. Review a description of the issue and understand how resolving it can improve your API.
Fixing API specification errors

The editor also supports AI error correction for issues in your OpenAPI definition that prevent processing. These errors appear inline in the editor and in the Problems tab. When available, click Apply to automatically resolve the issue using AI-generated fixes.

Requirements and availability

AI-powered recommendations currently apply only to OpenAPI specifications.

These AI enhancements and error features are available only when your application is connected to a valid API Management (APIM) instance. The features are enabled only if:
  • At least one configured APIM instance has the AI Assistant feature enabled.
  • A valid session with the APIM instance exists.

The recommendation engine runs in the backend using services provided by the connected APIM instance. IBM API Studio displays the results but does not perform any AI processing locally.