vault-report

Vault Knowledge Graph Report

Generate a structured knowledge graph analysis report for the entire vault in one go. Combine multiple query templates from obsidian-graph-query to produce four analysis modules + data summary.

Prerequisites

This skill depends on the obsidian-graph-query skill. Before execution:

Confirm that the obsidian-graph-query skill is installed (check whether the sibling ../obsidian-graph-query/SKILL.md exists)
Read ../obsidian-graph-query/references/vault-config.md to obtain the CLI path, vault name, and excluded folders
Query templates are in ../obsidian-graph-query/references/query-templates.md
The execution method follows the "Execution Mode" section of obsidian-graph-query (read template → substitute parameters → write temporary file → CLI eval)

Applicable Scenarios

"Analyze my knowledge base structure"
"Generate a knowledge graph report for my vault"
"Is my note library healthy?"
"Compare it with last year's analysis report"

Report Framework (4 Modules + Data Summary)

Module	Content	Main Data Sources
One: Vault Panoramic Overview	Scale, orphan rate, connectivity, folder distribution	vault-stats
Two: Activity Trends	Monthly output, peak activity months	vault-stats (monthlyCreation)
Three: Knowledge Focus and Structure	Hub Top 10, domain density, cross-domain links, orphan distribution	vault-stats + hubs + orphans-rich
Four: Knowledge Value and Risks	Most cited, outbound-only, Bridge risks	hubs + bridges + vault-stats
Data Summary	Structured tables (for visualization tools)	All of the above

Execution Flow

步驟 1. vault-stats（query-templates.md §8，無參數）
   → 模組一：totalNotes, totalLinks, avgLinksPerNote, orphanCount/Ratio,
            componentCount, largestComponent/Ratio, folderStats
   → 模組二：monthlyCreation
   → 模組三：crossFolderLinks/Ratio
   → 模組四：outOnlyCount/Notes

步驟 2. hubs(TOP_N=10)（query-templates.md §5）
   → 模組三：Hub 筆記 Top 10
   → 模組四：被引用最多的筆記（同一結果按 inDegree 排序）

步驟 3. hubs(FOLDER_FILTER=X) × 主要資料夾
   → 從步驟 1 的 folderStats 取筆記數前 5 大的資料夾
   → 每個資料夾跑一次 hubs(TOP_N=5, FOLDER_FILTER='資料夾/')
   → 模組三：各領域連結密度比較（密度 = 該領域總連結 / 該領域筆記數）

步驟 4. orphans-rich（query-templates.md §6，無 FOLDER_FILTER）
   → 模組三：孤島分布（按資料夾統計）

步驟 5. bridges（query-templates.md §4，無參數）
   → 模組四：articulationPoints 前 10 + bridgesByFolder

步驟 6. Agent 彙整上述 JSON → 產出 Markdown 報告

Steps 1-2 can be executed simultaneously (no dependency); Step 3 depends on the results of Step 1.

Report Output Format

# Vault 知識圖譜報告

> 產出時間：YYYY-MM-DD | Vault：<vault-name> | 筆記數：X

## 一、Vault 全景概覽

| 指標 | 數值 |
|------|------|
| 筆記總數 | X |
| 連結總數 | X |
| 平均每篇連結數 | X.XX |
| 孤島數量 | X（佔 XX.X%） |
| 連通分量數 | X |
| 最大分量涵蓋率 | XX.X% |

### 資料夾分布

| 資料夾 | 筆記數 | 連結數 | 孤島數 | 密度 |
|--------|--------|--------|--------|------|
| ... | ... | ... | ... | X.XX |

## 二、活躍度趨勢

| 月份 | 新增筆記數 |
|------|-----------|
| YYYY-MM | X |

活躍高峰月：YYYY-MM（X 篇）

## 三、知識重心與結構

### Hub 筆記 Top 10

| 排名 | 筆記 | 入度 | 出度 | 總連結 |
|------|------|------|------|--------|
| 1 | ... | ... | ... | ... |

### 各領域連結密度

| 領域 | 筆記數 | 連結數 | 密度 |
|------|--------|--------|------|
| ... | ... | ... | X.XX |

### 跨領域連結

跨資料夾連結數：X（佔總連結 XX.X%）

### 孤島分布

| 資料夾 | 孤島數 | 佔該資料夾 % |
|--------|--------|-------------|
| ... | ... | XX.X% |

## 四、知識價值與風險

### 被引用最多的筆記

（hubs 按 inDegree 排序前 10）

### 只出不進的筆記

outDegree > 0、inDegree = 0 的筆記共 X 篇

### Bridge 關鍵節點

（bridges articulationPoints 前 10，含 degree）

### Bridge 風險分布

| 資料夾 | Bridge 邊數 |
|--------|------------|
| ... | ... |

## 數據匯總

（前四模組的核心數據整理成 CSV-ready 表格，方便丟給視覺化工具）

Extended Usage: Cross-Validation with Qualitative Reports

After generating the report, it can be cross-validated with existing qualitative analysis reports (such as a NotebookLM annual review):

Test	Method
A: Cross-Domain Link Validation	vault-stats `crossFolderLinks` → path (§2) find cross-domain paths → LLM reads notes to determine semantic relationships
B: Hub Semantic Understanding	hubs Top 10 → LLM reads hub note content → explains the role it plays in the knowledge system
C: Bridge Risk Interpretation	bridges find key nodes → LLM reads bridge + neighbors on both sides → explains "what would happen if removed"
D: Orphan Association Reasoning	orphans-rich find orphans → neighbors (§1) find nearest connected → LLM determines "whether it should be connected"