compound-capture

To document verified fixes, extract module, symptoms, root cause, and prevention steps into structured YAML-frontmatter Markdown files. This ensures hard-won resolutions are validated against schema.yaml and stored consistently in a searchable knowledge base.

Consolidating feature branch artifacts requires discovering related artifacts, proposing project updates, and archiving them via automated commits. This auto-consolidation happens directly on the feature branch, optionally with user approval, to maintain a clean project state.

Yes, YAML frontmatter validation enforces enum categories and maps files to the correct knowledge-base paths. By validating against schema.yaml, the system ensures documentation is routed, categorized correctly, and cross-referenced with similar existing docs.

Yes, you can route sharp-edge insights back to skill, agent, or command definitions via interactive edits or optional GitHub issue creation. This ensures critical edge-case learnings are proposed directly to the relevant project files.

Headless mode works for capturing knowledge base entries without user interaction, but both headless and interactive modes are supported. Clear gating ensures missing context and YAML validation are handled appropriately across both operational modes.

Schema validation fails when capturing fixes if the YAML frontmatter does not match the required fields and enum categories defined in schema.yaml. The system enforces strict validation to ensure proper categorization and routing within the knowledge base.