llm-wiki · nashsu — Claude Code skill: query your local LLM Wiki
This skill wires your agent into the local LLM Wiki desktop app over its HTTP API: page search, reading, graph navigation and source rescan.
npx skills add nashsu/llm_wiki_skill --skill llm-wiki -a claude-code
npx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a claude-codeWhat llm-wiki does
The skill talks to a running LLM Wiki desktop app at 127.0.0.1:19828 through a plain JSON API. There is no client library to install — requests go out with whatever HTTP tool your environment already has, be it curl, fetch or requests. Pages live as wiki/**.md, raw documents sit under raw/sources/, and wikilinks tie them into a graph.
Auth is token-based. The token comes from the LLM_WIKI_API_TOKEN environment variable, from the app's Settings → API Server, or is skipped entirely when the unauthenticated mode is on. Check /api/v1/health first: if auth is not configured and the no-token mode is off, the user has to generate a token in Settings before anything works. Keep the token out of logs and URLs.
The skill fires only when the user names LLM Wiki outright, says "my wiki" or "my knowledge base", or asks to rescan sources. A generic "search my notes" request is left alone — that belongs to Obsidian, Notion and similar tools. Everything is read-only apart from the source rescan.
What it helps with
- find a page in your own knowledge base by query
- read a full wiki page by path
- browse the graph of links between pages
- rescan sources after editing files
- cite page paths back in the agent's answer
Example prompts for the agent
- what does my wiki say about rope embedding?
- read the page wiki/concepts/rope.md
- rescan my LLM Wiki sources
How to install llm-wiki
1Skills CLI (Vercel)
npx skills add nashsu/llm_wiki_skill --skill llm-wiki -a claude-codenpx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a claude-code2Manually: put the files in place
~/.claude/skills/nashsu-llm-wiki/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/nashsu-llm-wiki/
1Skills CLI (Vercel)
npx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a codex2Manually: put the files in place
.agents/skills/nashsu-llm-wiki/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a cursor2Manually: put the files in place
.cursor/skills/nashsu-llm-wiki/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install nashsu/llm_wiki_skill nashsu-llm-wiki2Manually: put the files in place
.github/skills/nashsu-llm-wiki/SKILL.md1Skills CLI (Vercel)
npx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a gemini-cli2Platform CLI
gemini skills install https://github.com/nashsu/llm_wiki_skillManually: put the files in place
.devin/skills/nashsu-llm-wiki/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add nashsu/llm_wiki_skill --skill nashsu-llm-wiki -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about llm-wiki
How do I install llm-wiki in Claude?
Run npx skills add nashsu/llm_wiki_skill --skill llm-wiki -a claude-code. The skill file then lands at ~/.claude/skills/nashsu-llm-wiki/SKILL.md.
What does llm-wiki do?
It searches wiki pages, reads their content, shows the link graph and triggers a source rescan inside the running LLM Wiki app. All of it goes through the local HTTP API.
What do I need before it works?
A running LLM Wiki app and a configured token — either in LLM_WIKI_API_TOKEN or under Settings → API Server. Hit /api/v1/health to see the current state.
Similar skills
All skills for Agent setup: memory, context, orchestration →llm-wiki by other authors comes first.