Wiki Query — Knowledge Retrieval

Before You Start

1. Read

   ~/.obsidian-wiki/config

   to get

   OBSIDIAN_VAULT_PATH

   (works from any project). Fall back to

   .env

   if you're inside the obsidian-wiki repo.
2. Read

   $OBSIDIAN_VAULT_PATH/index.md

   to understand the wiki's scope and structure

Retrieval Protocol

llm-wiki/SKILL.md

Step 1: Understand the Question

• Factual lookup — "What is X?" → Find the relevant page(s)
• Relationship query — "How does X relate to Y?" → Find both pages and their cross-references
• Synthesis query — "What's the current thinking on X?" → Find all pages that touch X, synthesize
• Gap query — "What don't I know about X?" → Find what's missing, check open questions sections

• Index-only mode — triggered by "quick answer", "just scan", "don't read the pages", "fast lookup". Stops at Step 3. Answers from frontmatter +

  index.md

  only.
• Normal mode — the full tiered pipeline below.

Step 2: Index Pass (cheap)

• You've already read

  index.md

  above — use it as the first filter. It lists every page with a one-line description and tags.
• Use

  Grep

  to scan page frontmatter only for title, tag, alias, and summary matches. A pattern like

  ^(title|tags|aliases|summary):

  scoped to vault

  .md

  files is far cheaper than content grep.
• Collect the top 5–10 candidate page paths ranked by:
  1. Exact title or alias match
  2. Tag match
  3. Summary field contains the query term

  4. index.md

     entry contains the query term

summary:

index.md

Step 2b: QMD Semantic Pass (optional — requires

QMD_WIKI_COLLECTION

in

.env

)

$QMD_WIKI_COLLECTION

No QMD? Skip to Step 3 and use

Grep

directly on the vault. QMD is faster and concept-aware but the grep path is fully functional. See

.env.example

for setup.

QMD_WIKI_COLLECTION

Grep

mcp__qmd__query:
  collection: <QMD_WIKI_COLLECTION>   # e.g. "knowledge-base-wiki"
  intent: <the user's question>
  searches:
    - type: lex    # keyword match — good for exact names, file paths, error messages
      query: <key terms>
    - type: vec    # semantic match — good for concepts, patterns, "what is X like"
      query: <question rephrased as a description>

papers

_raw/

QMD_PAPERS_COLLECTION

Step 3: Section Pass (medium cost — only if Steps 2/2b are inconclusive)

• Use

  Grep -A 10 -B 2 "<query-term>" <candidate-file>

  to get just the lines around the match.
• This usually returns 15–30 lines per hit instead of 100–500.
• If the section grep gives a clear answer, go straight to Step 5.

Step 4: Full Read (expensive — last resort)

• Read

  the top 3 candidates in full.
• Follow at most one hop of

  [[wikilinks]]

  from those pages if the answer requires cross-references.
• Check "Open Questions" sections for known gaps.
• If you're still short, then fall back to a broad content grep across the vault. Tell the user you escalated — this is the expensive path and they should know.

Step 5: Synthesize an Answer

• Cite specific wiki pages using

  [[page-name]]

  notation
• Note which step the answer came from ("found in summary" vs "grepped section" vs "full page read") — helps the user understand confidence
• If the wiki has contradictions, present both sides
• If the wiki doesn't cover something, say so explicitly
• Suggest which sources might fill the gap

Step 6: Log the Query

log.md

- [TIMESTAMP] QUERY query="the user's question" result_pages=N mode=normal|index_only escalated=true|false

Answer Format

Based on the wiki:

[Your synthesized answer with [[wikilinks]] to source pages]

Pages consulted: [[page-a]], [[page-b]], [[page-c]]

Gaps: [What the wiki doesn't cover that might be relevant]