gather-context

To quickly understand an unfamiliar codebase, you can scan its project identity, structure, commands, and dependencies to produce a concise contextual summary. This process reads files like README.md or SPEC.md to answer quick-start questions without creating any new files.

Mapping project structure and conventions involves enumerating top-level directories, entry points, and config files to infer coding patterns and key libraries. This approach reads documentation and configuration to deliver a chat-ready snapshot of the project layout.

Yes, you can scan a repository to identify dependencies and commands without creating files. The scanning process reads configuration files and README documentation to extract libraries, linters, and commands, delivering the entire summary directly in chat.

Reading README and SPEC files helps with onboarding by extracting project identity, framework details, and runtime requirements. This documentation provides the context needed to infer conventions and recent activity, resulting in a crisp project snapshot for quick onboarding.

Yes, this codebase summary approach works for identifying coding patterns and linters by reading configuration files. It infers conventions and key libraries directly from project setup files to inform a concise overview of the codebase's technical stack.

You should not use a chat-based codebase summary when you need persistent documentation files or detailed line-by-line code analysis. This approach provides a high-level contextual overview of structure and dependencies rather than generating persistent reference files.