A question-answering tool for an LLM Wiki, a curated personal knowledge base used as the source for responses.

Use it to search the wiki, read the most relevant pages, answer questions with citations, and optionally save a synthesis as a concept page.

It keeps answers grounded in the user's notes instead of relying only on general model knowledge, and reports when the wiki does not cover a topic.

wiki-query — Skill for Claude Code and Codex by bezata: A question-answering tool for an LLM Wiki, a curated…

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add skills/bezata/kobsidian/wiki-query

Any agent

npx skills add bezata/kObsidian --skill wiki-query

Clone the repo

git clone --depth 1 https://github.com/bezata/kObsidian

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/bezata/kobsidian/wiki-query.svg)](https://agentmods.dev/skills/bezata/kobsidian/wiki-query)

Your own site

<a href="https://agentmods.dev/skills/bezata/kobsidian/wiki-query"><img src="https://agentmods.dev/badge/skills/bezata/kobsidian/wiki-query.svg" alt="Measured on agentmods" height="20"></a>

Per session 74 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 798 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00074	$0.00798
Opus 5	$0.00037	$0.00399
Sonnet 5	$0.00015	$0.00160
Haiku 4.5	$0.00007	$0.00080

Measured 6d ago against content hash eb3a0bc1d7e7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

wiki-query scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/wiki-query/SKILL.md · 67 lines

How it starts

The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Wiki Query

You are answering from the user's curated wiki, not from training data. Treat wiki/ as source of truth. Always cite. If the wiki doesn't cover the question, say so and suggest an ingest.

When to use

User asks a factual / conceptual question and you know they have a wiki (either they just built one, or wiki.query returns results).
User says "per the wiki", "from our notes", "what do we have on X".
You're about to answer from general knowledge but suspect the wiki has a specific angle — check it first.

Steps

Call wiki.query with the topic. Use a concise, keyword-like topic (not a full question). limit defaults to 10 — fine for most queries; raise to 20 if the topic is broad.
Drill into the top matches. For the top 3-5 pages (ranked by score), call notes.read to get the full content. Prefer Concepts / Entities pages over raw Sources — they're already distilled.
Synthesize the answer. Write a short, direct answer in prose:
- Pull specific claims from the pages you read.
- Cite each claim with a wikilink: [[wiki/Concepts/memex.md|Memex]].
- If two pages disagree, surface the disagreement; don't paper over it. The wiki's contradictions are signal.
- If the wiki is silent on part of the question, say so explicitly — do not fall back to training data without flagging the gap.
Offer to compound. If your synthesis discovers a useful connection or produces a paragraph worth keeping, offer to file it back via wiki.summaryMerge into an existing concept page, or create a new one. Default: do NOT file automatically; ask first.
Log queries worth remembering. If the query took meaningful effort (you read 5+ pages, or produced a long synthesis), append a wiki.logAppend entry with op: "query" so the log reflects the wiki's real activity.

What not to do

Don't answer from training data and cite the wiki anyway. If you didn't read a page, don't wikilink it.
Don't just dump search results. wiki.query output is a map, not the answer. Read the pages, then synthesize.
Don't over-file. Not every query needs a permanent wiki page. File when the synthesis is non-obvious, or when the user asked you to.

Read the full file on GitHub · 67 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

6d ago First seen · 67 lines · 74 tokens per session scan A eb3a0bc1d7e7

Subscribe to this mod's changes

wiki-query is a skill published in the GitHub repository bezata/kObsidian (7 stars, last pushed 3d ago), licensed Apache-2.0. It adds 74 tokens to every session and 798 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Research codebase-indexing llm-wiki obsidian obsidian-mcp

wiki-query

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Wiki Query

When to use

Steps

What not to do

What this file has done since we first saw it

Other skills, from other repositories

llm-wiki

kb-lint

kb-index

ingest

lint

query