Quartz 5

Vault Lint Report — 2026-06-01

Properties1
tagsnote, organization, vault-health

3 min read

Summary

An internal vault health report generated by scripts/lint-vault.py (backlog threshold 14 days), tallying 43 must-fix issues, 33 warnings, and 106 informational items. The must-fix bucket is dominated by Clippings missing analyzed/source_hash frontmatter and one duplicate source_hash pair; warnings are almost entirely broken wikilinks in notes; and the informational tier is mostly “unknown tags” plus a set of orphan notes with no backlinks.

由 scripts/lint-vault.py 產生的內部 vault 健康報告（積壓門檻 14 天），統計出 43 項須修復、33 項須注意、106 項資訊。須修復多為 Clippings 缺少 analyzed／source_hash frontmatter，外加一組重複的 source_hash；須注意幾乎全是 notes 中的失效 wikilink；資訊層主要是「未知標籤」以及一批無反向連結的孤立筆記。

Key Points

  • 🔴 43 must-fix: mostly Clippings missing analyzed, source_hash (including all nine articles in this analysis batch), six robotics Clippings missing only analyzed, one note missing type, and a duplicate source_hash dc72c10aebb5c1fa shared by two near-identical “I stopped hitting Claude’s usage limits” files.
  • 🟡 33 warnings: all broken wikilinks, e.g. [[VaultBot]], [[robotics]], [[imitation-learning]], [[data]], plus several missing pasted-image and paper-title links across April–May notes.
  • 🟢 106 info: predominantly “unknown tags” not yet in the taxonomy (world-models, teleoperation, dataset, sim-to-real, etc.) and a list of orphan notes with no backlinks.
  • The orphan list flags potentially disconnected knowledge (e.g., claude-code-session-to-vps, robotic-dataset-survey, X-notes) that may need linking into maps/indexes.
  • Many broken wikilinks point to concept tags ([[robotics]], [[data]]) that appear intended as lowercase concept stubs but lack target notes.

Insights

The report is a snapshot of structural debt rather than content quality: the largest fixable category (missing analyzed/source_hash) is exactly what the analyze-vault workflow producing these AI files is meant to clear, so processing this batch should directly shrink the 🔴 count. The clustering of broken wikilinks around bare concept names suggests a systemic gap — concept stub notes (robotics, data, imitation-learning) are referenced but never created — which an index/map pass could resolve in bulk. The duplicate source_hash confirms a genuine duplicate-file pair worth deduping rather than a hashing error.

Connections

Raw Excerpt

🔴 需要修復（43 項）

  • Duplicate source_hash dc72c10aebb5c1fa: I stopped hitting Claude's usage limits - 10 things I changed.md, I stopped hitting Claude's usage limits - 10 things I changed 1.md

Graph View