How does it identify broken links?

The skill scans all markdown files for wiki-link patterns, verifies if the target file exists in the repository, and reports specific locations for any 'dangling' references.

What does the quick mode check?

Quick mode focuses on the three most critical health metrics: schema compliance (metadata), orphan note detection, and link integrity (dangling links).

What are Three-Space boundaries?

This is a structural rule that separates agent operations (ops), agent identity (self), and user knowledge (notes) to ensure search results and AI reasoning remain clean and relevant.

Where are the diagnostic reports stored?

Every run generates a detailed markdown report saved to 'ops/health/YYYY-MM-DD-report.md' for historical tracking of your vault's structural integrity.

Can this skill fix errors automatically?

The skill generates an actionable report with specific recommendations; users can then use Claude's editing capabilities to apply the suggested fixes or run maintenance commands.

Vault Health Diagnostics

Name: Vault Health Diagnostics
Author: agenticnotetaking

byagenticnotetaking

•

2,813

•

Productivity & Workflow

Runs comprehensive structural audits to ensure the integrity, link health, and schema compliance of markdown-based knowledge systems.

The Health skill is a specialized diagnostic engine for Claude Code designed to maintain high-quality decentralized knowledge vaults. It performs deep scans across eight critical dimensions—ranging from broken link detection and metadata schema validation to complex boundary checks between agent-generated content and user insights. By identifying orphaned notes, stale content, and structural inconsistencies, it provides a prioritized roadmap for vault maintenance. This skill is essential for users managing 'second brains' or complex documentation sets who need to prevent knowledge fragmentation and ensure their information architecture remains robust and searchable.

Key Features

01Actionable health reports with ranked fixes based on structural impact.

02Three-Space boundary enforcement to prevent agent-logic from polluting user notes.

03Detection of 'dangling links' with precise file and line-number reporting.

04Automated identification of stale notes and information restatements.

058-category diagnostic suite covering schema, links, orphans, and MOC coherence.

062,813 GitHub stars

Use Cases

01Identifying disconnected insights that need to be re-integrated into the knowledge graph.

02Performing weekly maintenance on a personal Obsidian or markdown-based second brain.

03Auditing structural integrity after a large batch of automated information processing.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add agenticnotetaking/arscontexta health

For use in Claude.ai and ChatGPT

Download Skill