onboard-project

You can generate a CLAUDE.md file by scanning config files and repository structures to detect technologies, map architecture, and extract codebase patterns. This produces a ready-to-use context briefing including CLAUDE.md and PROJECT_CONTEXT.md for AI context bootstrapping.

Analyzing codebase architecture for developer onboarding involves scanning repository structures to identify layouts like monorepo or monolith, exploring core elements such as routes and models, and mapping business domains. This surfaces conventions and clarifies responsibilities for new developers.

Pattern extraction works by exploring core code elements like entities, routes, and services to detect conventions such as naming and error handling. It groups logic into business domains or bounded contexts, clarifying responsibilities for developer onboarding.

Yes, you can update a pre-existing CLAUDE.md file during AI context bootstrapping. The process detects technologies from config files and integrates with projects needing an updated CLAUDE.md, generating a PROJECT_CONTEXT.md and a synthesis summary alongside it.

Yes, stack detection identifies technologies from config files like package.json and analyzes repository structure to determine monolith, monorepo, or other layouts. It supports codebase onboarding by mapping architecture and generating a project synthesis summary.

Avoid automated codebase onboarding when projects lack standard config files or recognizable repository structures, as stack detection and architecture analysis rely on scanning these elements. It is also less effective if core code elements like routes or models cannot be clearly mapped.