Zet query

Skill henrywen98/zettelkasten/skills/zet-query

This skill should be used when the user asks to "search knowledge base", "query notes", "zet query", "find in my notes", "what did I write about", "search my notes", "find notes about", "what do I know about", or wants to ask questions against their Zettelkasten knowledge base. Supports quick (MOC + index only) and deep (reads full notes) modes. Usage: /zet-query <question> [depth]. Depth: quick (MOC + grep only) | deep (reads full notes, default).From its SKILL.md

Install

npx -y skills add henrywen98/zettelkasten --skill zet-query

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.9 KB, as published. Nobody here has run it

Zettelkasten Query — Knowledge Base Q&A

Answer questions by navigating the Zettelkasten through MOCs and note links. This is the "thinking stage" — where the user engages with their accumulated knowledge.

Reference Specs

Frontmatter: ${CLAUDE_PLUGIN_ROOT}/references/frontmatter-spec.md
Vault structure: ${CLAUDE_PLUGIN_ROOT}/references/vault-structure.md

Workflow

1. Read MOC Index

Use Glob to list all MOC filenames in 2_maps/. Filter for MOCs whose filenames or titles are relevant to the question keywords, then read only those MOC files. This avoids reading all MOCs when there are many.

If 2_maps/ is empty, skip to step 2 and rely entirely on Grep.

2. Locate Relevant Notes

Use multiple strategies to find relevant content:

MOC navigation: Follow links from relevant MOCs to 1_zettel/ notes
Related Maps traversal: Check the ## Related Maps section of relevant MOCs to discover adjacent topic MOCs and their notes
Keyword search: Use Grep to search 1_zettel/ for question keywords and related terms
Link traversal: When reading a relevant note, follow its ## Links section to discover connected notes

Collect a set of candidate notes (aim for 5-20 depending on topic breadth).

3. Read Notes

Based on depth parameter:

quick: Read only the frontmatter (title, summary, tags) of candidate notes. Synthesize from summaries.
deep (default): Read full content of the most relevant candidate notes (up to ~15). Follow links to read additional notes if needed for completeness.

4. Synthesize Answer

Compose a structured answer that:

Directly addresses the question
Cites specific notes using wikilinks: [[note-title]]
Identifies gaps — topics the knowledge base doesn't cover
Suggests related notes the user might want to explore
Maintains the language of the question (Chinese question → Chinese answer)

5. Save Output

Save the answer to 3_output/YYYY-MM-DD-<slug>.md with frontmatter:

---
title: "Brief question summary"
query: "Original question text"
date: YYYY-MM-DD
sources: ["1_zettel/path/to/note1.md", "1_zettel/path/to/note2.md"]
---

Inform the user of the output file location.

6. Suggest Follow-ups

Based on the query results, suggest:

Related questions the user might want to explore
Notes that are tangentially related but might spark new connections
Gaps in the knowledge base that could be filled