content-enrich

Content Enrich

Collected transcripts and articles are often large blocks of raw text that no one wants to go back and read. This skill givesany collected output Automatically adds quickly digestible meta-information — it is a general-purpose capability of the 'enrichment layer', used together with all collection skills in the repository.

Environment Requirements

# 零 pip 依赖（仅 Python 标准库）
export DEEPSEEK_API_KEY="your-api-key"   # 必需，用于提炼

Usage

python scripts/enrich.py 某篇笔记.md              # 就地增强单篇
python scripts/enrich.py ./vault/外部输入/         # 批量增强整个目录下的 .md
python scripts/enrich.py 某篇.md --output ./out    # 不改原文件，输出到 out/
python scripts/enrich.py 某篇.md --force           # 已加工过的重新加工

Output

On the original Markdown (original content fully preserved):

frontmatter injection：summary(one-sentence summary),domain(domain),auto_tags(tag list),enriched: true
Insert at the top of the body ## 📝 摘要 block: one-sentence TL;DR, 3-5 key points, 'Worth a deep read?' judgment, tags

Idempotent: by default skips files that have already been enriched (enriched: true), can --force redo.

Why It Is Useful

Activates collected output: transcripts / articles let you see at a glance what they are about and whether they are worth a deep read
Facilitates knowledge base aggregation：auto_tags / domain into the frontmatter, Obsidian Dataview can directly aggregate by tag / domain
Invest once, benefit everywhere: Douyin / Bilibili / Xiaohongshu / X / WeChat Official Accounts …… any collected output can be enriched

Workflow Integration

采集（任意 *-ingest / *-transcribe skill）
  → content-enrich（补摘要 + 要点 + 标签）
  → knowledge-base-management 入库（按 auto_tags / domain 聚合检索）

Reference

DeepSeek — used for LLM-based extraction