tachikoma-schema

Schema analysis examines your workspace structure and proposes improvements by sampling entity types, identifying naming patterns, and generating consolidated decisions. Run tachikoma-schema to scan your workspace, read schema.yaml and tachikoma-summary.yaml, then create decision files documenting recommended schema updates.

Schema cleanup involves inspecting your existing schema.yaml, sampling markdown files to identify entity type patterns, proposing naming consistency improvements, and recording findings in decisions/. The process creates traceability by updating tachikoma-summary.yaml with analysis results.

Yes. Schema analysis is designed for creating or updating schema.yaml by first identifying entity types across your workspace. It samples multiple markdown files per candidate type, detects naming patterns, and proposes entity type definitions before you finalize schema changes.

Listing and analyzing workspace root directories reveals structural patterns that inform entity type definitions. Schema analysis uses these patterns alongside markdown sampling to propose entities that align with your actual workspace organization.

Tachikoma-summary.yaml records findings from workspace analysis, providing traceability for schema improvement decisions. After schema analysis generates decision files, the summary is updated to document what was proposed and why.