temple-generator

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Temple Generator

Generate a 3D interactive knowledge visualization from any Obsidian vault. The output is a single HTML file (Three.js) with concentric entity rings, audio, discovery mechanics, and multi-scale semantic zoom.

从任意Obsidian知识库生成3D交互式知识可视化内容。输出为单个HTML文件（基于Three.js），包含同心实体环、音频、探索机制以及多尺度语义缩放功能。

When to Use

适用场景

User wants to visualize any Obsidian vault as a 3D knowledge map
User wants to compare two vaults/document sets visually
User wants to regenerate the temple from scratch with fresh vault analysis

用户希望将任意Obsidian知识库可视化为3D知识图谱
用户希望直观对比两个知识库/文档集
用户希望基于全新的知识库分析结果重新生成圣殿

Architecture

架构

Two-part system:

Generation pipeline (this skill): discovers structure, names it, scores confidence, exports a scene package
Runtime renderer (template): handles navigation, transitions, audio, discovery

Pre-generate meaning. Runtime-render experience.

分为两部分的系统：

生成流水线（本技能）：发现结构、命名、置信度评分、导出场景包
运行时渲染器（模板）：处理导航、过渡、音频、探索功能

预先生成语义信息，运行时渲染交互体验。

Workflow

工作流程

Step 1: Scan the Vault

步骤1：扫描知识库

Run

python3 ~/.claude/skills/temple-generator/scripts/extract_entities.py <vault_path>

This produces

vault-scan.json

with:

Files: path, title, tags, outgoing links, backlink counts, word count, folder, frontmatter
Graph: adjacency list with bidirectional link counts
Centrality: degree centrality per node
Clusters: detected groups of tightly linked notes

运行

python3 ~/.claude/skills/temple-generator/scripts/extract_entities.py <vault_path>

。

此命令会生成

vault-scan.json

，包含：

文件：路径、标题、标签、出站链接、反向链接计数、字数、文件夹、前置元数据
图谱：带双向链接计数的邻接表
中心性：每个节点的度中心性
集群：检测到的紧密关联笔记组

Step 2: Read the Scan + Sample Notes

步骤2：读取扫描结果与示例笔记

Read
```
vault-scan.json
```
Read the top ~20 nodes by centrality (first 100 lines each)
Read
```
references/classification-guide.md
```
for entity type heuristics
Read 3-5 representative notes to calibrate the vault's "voice" (formal/informal, domain jargon, language)

读取
```
vault-scan.json
```
读取按中心性排序的前约20个节点（每个节点读取前100行）
读取
```
references/classification-guide.md
```
获取实体类型启发规则
读取3-5篇代表性笔记，校准知识库的“语气”（正式/非正式、领域术语、语言风格）

Step 3: Classify Entities

步骤3：实体分类

Using

references/classification-guide.md

, assign each significant node to an entity type. Maintain two vocabularies:

canonical: neutral labels for portability (
```
anxiety-management
```
,
```
fermentation-process
```
)
poetic: mythic/art labels for the installation (
```
The Ferment Gate
```
,
```
The Cortisol Throne
```
)

Target counts per type (adjust for vault size):

Type	Small vault (< 100)	Medium (100-500)	Large (500+)
Gods	2-3	3-5	5-7
Demigods	3-7	5-12	8-15
Tensions	2-4	3-7	5-9
Narratives	2-5	5-10	8-12
Blind spots	1-3	3-5	4-7
Spirits	1-3	3-5	3-5
Research	5-15	10-25	15-30
Values	2-5	3-8	5-10
Trails	2-5	3-8	5-10
Questions	3-6	5-10	8-12
Depths	2-5	5-10	8-15
Crystals	1-3	2-5	3-6

使用

references/classification-guide.md

，为每个重要节点分配实体类型。维护两种词汇体系：

标准词汇：用于可移植性的中性标签（如
```
anxiety-management
```
、
```
fermentation-process
```
）
诗意词汇：用于可视化场景的神话/艺术化标签（如
```
The Ferment Gate
```
、
```
The Cortisol Throne
```
）

各类型目标数量（根据知识库大小调整）：

类型	小型知识库（<100篇）	中型知识库（100-500篇）	大型知识库（500+篇）
神祇	2-3	3-5	5-7
半神	3-7	5-12	8-15
张力	2-4	3-7	5-9
叙事	2-5	5-10	8-12
盲区	1-3	3-5	4-7
灵体	1-3	3-5	3-5
研究	5-15	10-25	15-30
价值观	2-5	3-8	5-10
路径	2-5	3-8	5-10
问题	3-6	5-10	8-12
深渊	2-5	5-10	8-15
晶体	1-3	2-5	3-6

Step 4: Build Abstraction Levels

步骤4：构建抽象层级

Levels are confidence-gated — only include a level if the vault supports it.

Level 0 — Entities (always exists): individual nodes with positions, connections, descriptions.

Level 1 — Domains (requires >= 3 meaningful clusters): groups of related entities. Each domain has:

```
canonical
```
+
```
poetic
```
name
member entity keys
centroid position (weighted average of member positions)
representative exemplar (most central member)
description (1-2 sentences in vault voice)
confidence score (0-1)

Level 2 — Axes (requires >= 2 interpretable opposing pairs): fundamental tensions. Each axis has:

two poles with names and descriptions
member domains per pole
axis description
confidence score

Level 3 — Comparison (requires two vaults + sufficient alignment): shared/unique analysis.

Read

references/merge-algorithm.md

for dual-graph logic.

层级采用置信度门控——仅当知识库支持时才包含对应层级。

层级0 — 实体（始终存在）：包含位置、连接关系、描述的单个节点。

层级1 — 领域（需要≥3个有意义的集群）：相关实体的分组。每个领域包含：

标准名称 + 诗意名称
成员实体键
质心位置（成员位置的加权平均值）
代表性范例（最核心的成员）
描述（1-2句，匹配知识库语气）
置信度评分（0-1）

层级2 — 轴（需要≥2个可解释的对立对）：核心张力。每个轴包含：

两个带有名称和描述的极点
每个极点对应的领域成员
轴描述
置信度评分

层级3 — 对比（需要两个知识库 + 足够的匹配度）：共享/独有分析结果。

阅读

references/merge-algorithm.md

获取双图逻辑。

Step 5: Generate Scene Package

步骤5：生成场景包

Follow the schema in

references/entity-schema.md

to produce

temple-data.json

Include:

```
entities
```
: all classified nodes
```
levels
```
: abstraction layers with zoom thresholds
```
mappings
```
: entity → domain → axis crosswalks
```
comparison
```
: (if dual-graph) shared/unique/alignment data
```
audio
```
: motif hints per type and level
```
style
```
: poetic vocabulary, intro text, color palette, layer definitions
```
confidence
```
: per-abstraction and per-alignment scores

遵循

references/entity-schema.md

中的 schema 生成

temple-data.json

。

包含内容：

```
entities
```
：所有已分类的节点
```
levels
```
：带有缩放阈值的抽象层级
```
mappings
```
：实体→领域→轴的交叉映射
```
comparison
```
：（如果是双图模式）共享/独有/匹配度数据
```
audio
```
：各类型和层级的音频主题提示
```
style
```
：诗意词汇、介绍文本、调色板、层定义
```
confidence
```
：各抽象层级和匹配度的评分

Step 6: Generate HTML

步骤6：生成HTML

Copy

~/.claude/skills/temple-generator/assets/temple-template.html

to the output location

If
```
--inline
```
flag: embed the JSON data as
```
const TEMPLE_DATA = {...};
```
inside the HTML
Otherwise: place
```
temple-data.json
```
alongside the HTML

将

~/.claude/skills/temple-generator/assets/temple-template.html

复制到输出位置

如果使用
```
--inline
```
参数：将JSON数据以
```
const TEMPLE_DATA = {...};
```
的形式嵌入HTML中
否则：将
```
temple-data.json
```
放在HTML文件同一目录下

Step 7: Report

步骤7：生成报告

Show the user:

Entity counts by type
Abstraction levels generated (with confidence scores)
Top 5 gods/central entities
Detected tensions
If dual-graph: overlap percentage and shared domains

向用户展示：

各类型实体的数量
生成的抽象层级（含置信度评分）
前5个核心实体（神祇）
检测到的张力
如果是双图模式：重叠百分比和共享领域

Dual-Graph Mode

双图模式

When

--compare vault_path_2

is provided:

Scan both vaults independently (Step 1)
Classify entities for each vault (Steps 2-3)
Run merge algorithm from
```
references/merge-algorithm.md
```
Generate merged scene package with source attribution
Template renders shared scaffold with divergence offsets

当提供

--compare vault_path_2

参数时：

独立扫描两个知识库（步骤1）
为每个知识库分类实体（步骤2-3）
运行
```
references/merge-algorithm.md
```
中的合并算法
生成带有来源标注的合并场景包
模板渲染共享框架并显示差异偏移

Quality Guidelines

质量指南

Skip trivial notes (daily todos, admin logs, empty stubs)
Prefer nodes that reveal the vault's actual concerns, not its filing system
Write in the vault's own voice, calibrated from sample notes
If a level lacks confidence, omit it rather than fabricating structure
Each abstraction level must be backed by membership weights, exemplars, and provenance
"The abstraction hierarchy should be semantic, not just geometric"

跳过无关紧要的笔记（每日待办、管理日志、空草稿）
优先选择能体现知识库核心关注点的节点，而非其文件系统结构
使用与知识库一致的语气，基于示例笔记校准
如果某层级置信度不足，应省略而非编造结构
每个抽象层级必须有成员权重、范例和来源作为支撑
"抽象层级应基于语义，而非仅几何结构"

Audio Guidance for Template

模板音频指南

The template's audio system should respect hierarchical continuity across zoom levels:

L0 (close): localized, identity-rich — entity whispers and textures
L1 (medium): regional harmonic beds, cluster pulses
L2 (far): sparse drones, tension-based tonal movement
L3 (comparison): stereo/dialogic between two vault voices

Zoom should feel like changing resolution, not changing universes. Motifs relate across scales.

模板的音频系统应在缩放层级间保持层级连续性：

L0（近距离）：本地化、富有辨识度的实体低语和纹理音效
L1（中距离）：区域性和声背景、集群脉冲音效
L2（远距离）：稀疏的低音 drone、基于张力的音调变化
L3（对比模式）：两个知识库声音的立体声/对话式音效

缩放应像是改变分辨率，而非切换宇宙。主题在不同尺度间保持关联。