feature-domain-expert

You preserve bounded-context domain knowledge for SDD by authoring feature-level files that separate long-lived business rules from per-change delta specs. This keeps domain knowledge stable across propose, spec, design, and verify phases.

The canonical structure for feature domain knowledge files requires a six-section format in ai-context/features/feature-domain-expert.md. It must include a frontmatter header with name and description to enable discovery and tooling.

You separate domain knowledge from per-change delta specs by maintaining stable feature-level files in ai-context/features/. This approach stores business rules, invariants, and integration points independently from individual change specifications.

You need to update feature-level domain knowledge files when creating or modifying bounded-context features to guide SDD phases. This ensures long-lasting rules, integration points, and known gotchas remain accurate across changes.

Yes, the feature-domain-expert file format requires a frontmatter header containing name and description for discovery and tooling. This header accompanies the canonical six-section structure within the ai-context/features/ directory.