onboard

Automating repository onboarding requires scanning existing source code and dependency manifests to draft essential configuration files and workspace documentation. This process generates configuration surfaces and domain experts based on actual project structure rather than assumptions.

Automated onboarding analyzes project structure by surveying source code, documentation, and configuration files via local filesystem read access. It infers conventions and domains directly from dependency manifests, producing evidence-based configuration drafts for human verification.

Yes, you need local filesystem read access to survey source code, documentation, and project configuration files. This access is required to analyze existing repository structures and dependency manifests for drafting configuration surfaces and domain experts.

The onboarding process only generates drafts for human verification, ensuring alignment with existing project standards. It produces workspace maps and configuration files as review-first outputs without modifying existing source code or repository files.

Scaffolding workspace documentation generates workspace-CLAUDE.md, .aif/config.yml, and domain-specific experts. These configuration surfaces are drafted based on evidence from the repository structure and dependency manifests.

Automated scaffolding is suited for initial toolkit adoption scenarios where manual setup of workspace documentation and config files is required. It reduces friction by drafting configuration surfaces and domain experts based on actual codebase analysis.