zego-extend-claude-standards

You collect repository context for CLAUDE.local.md through structured interviews that gather detailed information on purpose, ownership, interfaces, and dependencies. The skill integrates this interview data directly into the documentation file.

Interview-driven documentation is a process that collects repository context by interviewing engineers. It classifies information across canonical topics as present, thin, or absent, presenting a gap summary to guide structured updates.

You identify missing information by classifying each documentation topic as present, thin, or absent. The skill evaluates your existing context and presents a gap summary to highlight areas requiring further engineering input.

Yes, interview-driven context collection works without existing documentation. The skill classifies canonical topics as absent when no prior context exists, using the interview process to build the repository documentation from scratch.

The best way to deepen repository context is using a collaborative update process. The skill merges information collected through structured interviews with multiple engineers directly into the CLAUDE.local.md file.

Your repository context lacks ownership and dependency details because those canonical topics are likely classified as thin or absent. A structured interview process collects this missing information and merges it into your documentation.