| name | wiki-query |
| description | Answer questions using the Obsidian wiki vault. Reads hot cache first, then index, then relevant pages. Synthesizes answers with citations. Files good answers back as wiki pages. Supports quick, standard, and deep modes. Triggers on: what do you know about, query:, what is, explain, summarize, find in wiki, search the wiki, based on the wiki, wiki query quick, wiki query deep. |
wiki-query: Query the Wiki
The wiki has already done the synthesis work. Read strategically, answer precisely, and file good answers back so the knowledge compounds.
Default answers are in Simplified Chinese unless the user explicitly asks for another language. Preserve page titles, repository names, API names, and conventional technical terms as written in the wiki.
Transport (v1.7+)
Reads should prefer the same transport the rest of the plugin uses. Consult .vault-meta/transport.json (auto-created by bash scripts/detect-transport.sh) and use the preferred entry:
- cli —
obsidian-cli read "$VAULT" "$NOTE" and obsidian-cli search "$VAULT" "<query>" (Obsidian-native ranking); see skills/wiki-cli/SKILL.md
- mcp-obsidian / mcpvault —
mcp__obsidian-vault__read_note, search_notes; see skills/wiki/references/mcp-setup.md
- filesystem — Codex's
Read and Glob/Grep tools (final floor; always works)
Full decision tree: wiki/references/transport-fallback.md. Quick mode (hot.md only) is transport-agnostic — always uses Read.
Retrieval (v1.7+)
If wiki-retrieve is feature-detected — [ -x scripts/retrieve.py ] && [ -d .vault-meta/chunks ] && [ -f .vault-meta/bm25/index.json ] — Standard and Deep modes consult it BEFORE the legacy hot→index→drill chain:
python3 scripts/retrieve.py "<the user's question verbatim>" --top 5
Output is JSON with a candidates array. Each candidate has absolute_path to the source page, a snippet, and bm25_score + rerank_score. Read the cited pages (using the transport selector from §Transport above) and synthesize with chunk-level citation.
If retrieve.py exits 10 (feature not provisioned), or any step in the pipeline errors, fall back to the v1.6 legacy read order described in the Standard/Deep workflows below — no user-visible breakage.
Quick mode always skips retrieval (hot.md only — keeps the ~1,500 token budget intact).
Full spec: skills/wiki-retrieve/SKILL.md. Setup: bash bin/setup-retrieve.sh. The legacy read-order workflows below remain authoritative when wiki-retrieve is not installed.
Query Modes
Three depths. Choose based on the question complexity.
| Mode | Trigger | Reads | Token cost | Best for |
|---|
| Quick | query quick: ... or simple factual Q | hot.md + index.md only | ~1,500 | "What is X?", date lookups, quick facts |
| Standard | default (no flag) | hot.md + index + 3-5 pages | ~3,000 | Most questions |
| Deep | query deep: ... or "thorough", "comprehensive" | Full wiki + optional web | ~8,000+ | "Compare A vs B across everything", synthesis, gap analysis |
Quick Mode
Use when the answer is likely in the hot cache or index summary.
- Read
wiki/hot.md. If it answers the question, respond immediately.
- If not, read
wiki/index.md. Scan descriptions for the answer.
- If found in index summary, respond and do not open any pages.
- If not found, say "Not in quick cache. Run as standard query?"
Do not open individual wiki pages in quick mode.
Standard Query Workflow
- Read
wiki/hot.md first. It may already have the answer or directly relevant context.
- Read
wiki/index.md to find the most relevant pages (scan for titles and descriptions).
- Read those pages. Follow wikilinks to depth-2 for key entities. No deeper.
- Synthesize the answer in chat in Simplified Chinese by default. Cite sources with wikilinks:
（来源：[[Page Name]]）.
- Offer to file the answer in Chinese: "这个分析值得保留。要保存为
wiki/questions/answer-name.md 吗？"
- If the question reveals a gap: say "当前 wiki 还不足以回答 X。要不要找一个来源补上？"
Deep Mode
Use for synthesis questions, comparisons, or "tell me everything about X."
- Read
wiki/hot.md and wiki/index.md.
- Identify all relevant sections (concepts, entities, sources, comparisons).
- Read every relevant page. No skipping.
- If wiki coverage is thin, offer to supplement with web search.
- Synthesize a comprehensive answer with full citations.
- Always file the result back as a wiki page. Deep answers are too valuable to lose.
Token Discipline
Read the minimum needed:
| Start with | Cost (approx) | When to stop |
|---|
| hot.md | ~500 tokens | If it has the answer |
| index.md | ~1000 tokens | If you can identify 3-5 relevant pages |
| 3-5 wiki pages | ~300 tokens each | Usually sufficient |
| 10+ wiki pages | expensive | Only for synthesis across the entire wiki |
If hot.md has the answer, respond without reading further.
Index Format Reference
The master index (wiki/index.md) looks like:
## Domains
- [[Domain Name]]: description (N sources)
## Entities
- [[Entity Name]]: role (first: [[Source]])
## Concepts
- [[Concept Name]]: definition (status: developing)
## Sources
- [[Source Title]]: author, date, type
## Questions
- [[Question Title]]: answer summary
Scan the section headers first to determine which sections to read.
Domain Sub-Index Format
Each domain folder has a _index.md for focused lookups:
---
type: meta
title: "Entities Index"
updated: YYYY-MM-DD
---
# Entities
## People
- [[Person Name]]: role, org
## Organizations
- [[Org Name]]: what they do
## Products
- [[Product Name]]: category
Use sub-indexes when the question is scoped to one domain. Avoid reading the full master index for narrow queries.
Filing Answers Back
Good answers compound into the wiki. Don't let insights disappear into chat history.
When filing an answer:
---
type: question
title: "Short descriptive title"
question: "The exact query as asked."
answer_quality: solid
created: YYYY-MM-DD
updated: YYYY-MM-DD
tags: [question, <domain>]
related:
- "[[Page referenced in answer]]"
sources:
- "[[wiki/sources/relevant-source.md]]"
status: developing
language: zh
---
Then write the answer as the page body. Include citations. Link every mentioned concept or entity.
After filing, add an entry to wiki/index.md under Questions and append to wiki/log.md.
Gap Handling
If the question cannot be answered from the wiki:
- Say clearly: "I don't have enough in the wiki to answer this well."
- Identify the specific gap: "I have nothing on [subtopic]."
- Suggest: "Want to find a source on this? I can help you search or process one."
- Do not fabricate. Do not answer from training data if the question is about the specific domain in this wiki.
How to think (10-principle mapping)
When working on this skill, apply the 10-principle loop. See skills/think/SKILL.md for the canonical framework.
| # | Principle | Application here |
|---|
| 1 | OBSERVE (ext) | Read wiki/hot.md first, then wiki/index.md, then specific pages. Don't skip the cache. |
| 2 | OBSERVE (int) | Am I synthesizing from training-data memory when I should be citing wiki pages? Check the source of each claim. |
| 3 | LISTEN | What is the user's REAL question? The surface query is often a proxy for a deeper need. |
| 4 | THINK | Quick / standard / deep mode? Match depth to question complexity, not eagerness. |
| 5 | CONNECT (lat) | Are there pages I missed that would CHANGE the answer? Cross-check related pages before answering. |
| 6 | CONNECT (sys) | Hot cache + index + wiki-retrieve (when provisioned) layer into a single retrieval pipeline. |
| 7 | FEEL | Cite specific pages, not vague references. Future-me wants traceability back to the source page. |
| 8 | ACCEPT | When the wiki doesn't have the answer, say so explicitly. Don't fabricate from training data. |
| 9 | CREATE | The answer with citations + an offer to file the answer if it's worth keeping. |
| 10 | GROW | Questions the wiki can't answer are content gaps — log them as autoresearch inputs. |
