What problem does it solve?
Identify and fix quality issues in knowledge base content by validating structure, required fields, and link integrity, while ensuring abstracts stay aligned with full content. It runs as a non-blocking gate and reports warnings to prevent workflow disruption.
Core Features & Use Cases
- Structural checks: YAML frontmatter presence; parseable; required fields exist (type, title, date); Markdown structure validity; file naming conventions.
- Schema checks: Validate that related files (e.g., context.md, signal.md, claims) meet required schema fields (e.g., context sections, date, priority, status, tags, etc.).
- Link checks: Validate internal references exist; anchors exist; avoid circular references beyond depth 3.
- Fidelity checks: Read L0 abstracts and full content; compute semantic alignment; flag drift when alignment falls below a threshold (e.g., 0.7).
- Duplicate detection: Detect high-content overlap across files and flag for deduplication.
Quick Start
Run the /verify command across your knowledge base to initiate quality validation.