rai-discover

Yes, you can build a knowledge graph for multi-language codebases by running automated language detection and symbol extraction. The pipeline processes a project root directory and generates discovery artifacts to map the structure across different programming languages.

Mapping codebase structure generates discovery artifacts including context.yaml, components-validated.json, module documentation, and system documentation. These outputs capture the validated components and architectural context needed for project onboarding and session workflows.

To generate module and system documentation scaffolding from an existing codebase, you run the discovery pipeline on your project root. It automatically extracts symbols and outputs structured module docs and system docs alongside the validated components file.

No, the codebase discovery pipeline does not require manual language configuration because it uses automated language detection. You provide a project root input and the pipeline optionally detects languages to process multi-language codebases automatically.

The best way to validate components for software project onboarding is to run automated discovery to generate a components-validated.json file. This artifact captures the validated structural components and integrates with session workflows to speed up project setup.