← 返回 Skills 市场
hal-9909

LLM-WikiMind MCP Setup

作者 HAL-9909 · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
76
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install llm-wikimind-mcp
功能描述
Install and configure the LLM-WikiMind MCP server — a local knowledge base built on Karpathy's LLM Wiki pattern. Triggers: install wikimind, setup knowledge...
使用说明 (SKILL.md)

LLM-WikiMind MCP Setup Skill

Set up the LLM-WikiMind MCP server — a local knowledge base built on Karpathy's LLM Wiki pattern.

What you get: 5 MCP tools (wiki_search, wiki_get, wiki_list, wiki_ingest_note, wiki_domains) backed by BM25 full-text search over your local Markdown files. No embeddings, no vector DB, no cloud.

Trigger phrases

  • "Install WikiMind" / "Set up my knowledge base MCP"
  • "Configure wiki-kb" / "llm-wikimind setup"
  • "安装知识库 MCP" / "配置 wiki-kb" / "搭建个人知识库"

Installation Workflow

Step 1: Install the only dependency

pip3 install qmd

Step 2: Clone the repo

git clone https://github.com/HAL-9909/llm-wikimind
cd llm-wikimind

Step 3: Initialize the wiki

Run the interactive setup — it will ask where to store the knowledge base and handle everything else:

./wikimind init

Options:

Command When to use
./wikimind init Create a fresh wiki (interactive, asks for path)
./wikimind init ~/my-notes --adopt Adopt an existing Markdown directory

init automatically:

  • Creates the standard wiki directory structure
  • Copies the MCP server and watcher into the wiki
  • Builds the initial BM25 search index
  • Prints the exact config snippet to paste into the AI client

Step 4: Register the MCP server

The init command prints the exact snippet. For reference:

CatDesk / OpenClaw:

catdesk mcp add --name wiki-kb --json '{
  "command": "python3",
  "args": ["\x3CWIKIMIND_ROOT>/.wiki-mcp/server.py"],
  "env": {"WIKIMIND_ROOT": "\x3CWIKIMIND_ROOT>"}
}'

Replace \x3CWIKIMIND_ROOT> with the path chosen in Step 3 (e.g. ~/Documents/wiki).

Claude Desktop — add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "wiki-kb": {
      "command": "python3",
      "args": ["\x3CWIKIMIND_ROOT>/.wiki-mcp/server.py"],
      "env": { "WIKIMIND_ROOT": "\x3CWIKIMIND_ROOT>" }
    }
  }
}

Step 5: Start the auto-sync watcher

./wikimind start

Auto-start on login:

echo '/path/to/llm-wikimind/wikimind start > /dev/null 2>&1' >> ~/.zshrc

Step 6: Verify

./wikimind status

Expected output:

→ Wiki root:  ~/Documents/wiki
✓ Wiki exists — 1 domain(s), 3 pages
✓ Watcher running (pid 12345)
✓ qmd installed

MCP Tools Reference

Once registered, these 5 tools are available to any MCP-compatible AI client:

Tool Description
wiki_search BM25 full-text search across all domains
wiki_get Read a specific page in full
wiki_list List pages by domain and/or type
wiki_ingest_note Write a new page + rebuild index + sync cache
wiki_domains List all registered domains and their trigger keywords

Auto-update: always in sync

The watcher runs in the background and keeps everything current automatically:

  • Add a new domain folder → detected within 10 seconds → MCP tool descriptions updated
  • Edit DOMAIN.md keywords → AI knows about the change in the next conversation
  • AI writes a new page via wiki_ingest_note → index rebuilt automatically

No restarts. No manual config changes.


Adding knowledge after setup

Install the companion ingest skill:

npx clawhub@latest install wikimind-ingest

Then just say: "Add this to my knowledge base: [paste content or URL]"


Troubleshooting

qmd: command not found → Run pip3 install qmd

MCP server not responding → Check ./wikimind status and ensure WIKIMIND_ROOT is set correctly

Watcher not running → Run ./wikimind start; add to ~/.zshrc for auto-start

New domain not detected → Ensure DOMAIN.md has a keywords frontmatter field; check ./wikimind status

安全使用建议
This skill appears to be a legitimate installer for a local knowledge-base MCP, but take a few precautions before proceeding: - Inspect the GitHub repo code (especially .wiki-mcp/server.py and the 'wikimind' scripts) before running them; cloned code runs locally and can do anything your user account can do. - Check the PyPI package 'qmd' (pip3 install qmd) before installing; consider using a Python virtualenv to avoid changing your global site-packages. - Ensure you have pip3 available (the SKILL.md uses pip3, but required binaries omit it); install pip3 or adapt commands accordingly. - Back up ~/.zshrc and any client config files (e.g., Claude config) before letting the installer append or edit them. - Be aware that the watcher starts a background process and the server will be reachable locally; if you care about network exposure, review the server binding/ports and use a firewall or x-local-only binding. - If you want stronger isolation, run the service under a dedicated user or inside a container. - Ask the publisher to update metadata to declare pip3 as a required binary and to mention WIKIMIND_ROOT as an expected env variable so the manifest matches the runtime instructions.
功能分析
Type: OpenClaw Skill Name: llm-wikimind-mcp Version: 1.0.0 The skill instructions in SKILL.md require the agent to clone an external GitHub repository (HAL-9909/llm-wikimind), execute local scripts (./wikimind), and establish persistence by modifying the user's shell configuration (~/.zshrc). While these actions are consistent with the stated purpose of installing a local knowledge base, the execution of unverified external code and the modification of startup files are high-risk behaviors that could be used for malicious purposes.
能力评估
Purpose & Capability
The name/description (LLM-WikiMind MCP server) aligns with the required binaries (python3, git) and the runtime actions (git clone, start a Python server). However the SKILL.md uses pip3 but pip3 is not listed in required binaries, and it expects the WIKIMIND_ROOT environment variable in the registration examples even though requires.env is empty. Those omissions are inconsistent with the declared metadata.
Instruction Scope
Instructions are focused on installing, initializing, and running the local wiki and MCP server. They do instruct modifying user configuration files (append auto-start line to ~/.zshrc and editing client config like Claude's JSON) and starting a background watcher that auto-syncs. The instructions do not ask for secrets or for reading unrelated system files, but they do direct writes to user config files and create a persistent background process.
Install Mechanism
Installation is instruction-only: git clone from GitHub (expected) and pip3 install qmd (PyPI). These are common, but installing arbitrary PyPI packages and running code from a cloned repo can execute arbitrary code — the package 'qmd' and the repository contents should be reviewed before running. The optional npx clawhub@latest command (npm) is another package install to be reviewed if used.
Credentials
No credentials or sensitive env vars are requested, which fits a local-only wiki. However the SKILL.md expects WIKIMIND_ROOT to be set/provided when registering the MCP server, yet requires.env is empty. Also pip3 is used but not declared as a required binary. These are mismatches between declared requirements and the runtime instructions.
Persistence & Privilege
The skill does not force inclusion (always: false). Still, it recommends persistent changes: starting a background watcher, appending a command to ~/.zshrc for auto-start, and registering a local MCP server in client configs. Those are normal for a local service installer but are persistent and will run code on login/in the background until removed.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install llm-wikimind-mcp
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /llm-wikimind-mcp 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
llm-wikimind-mcp 1.0.0 initial release - Provides step-by-step setup for a local LLM knowledge base (WikiMind MCP) using Karpathy's LLM Wiki pattern. - Enables five MCP tools for knowledge management: wiki_search, wiki_get, wiki_list, wiki_ingest_note, and wiki_domains. - Uses BM25 full-text search across Markdown files; no embeddings or cloud required. - Includes auto-sync watcher that keeps the knowledge base and tool list up-to-date. - Detailed instructions for installation, configuration, troubleshooting, and MCP client registration.
元数据
Slug llm-wikimind-mcp
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

LLM-WikiMind MCP Setup 是什么?

Install and configure the LLM-WikiMind MCP server — a local knowledge base built on Karpathy's LLM Wiki pattern. Triggers: install wikimind, setup knowledge... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 76 次。

如何安装 LLM-WikiMind MCP Setup?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install llm-wikimind-mcp」即可一键安装,无需额外配置。

LLM-WikiMind MCP Setup 是免费的吗?

是的,LLM-WikiMind MCP Setup 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

LLM-WikiMind MCP Setup 支持哪些平台?

LLM-WikiMind MCP Setup 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 LLM-WikiMind MCP Setup?

由 HAL-9909(@hal-9909)开发并维护,当前版本 v1.0.0。

💬 留言讨论