# gknowledge MCP server

> Allows an AI client to search, read, and grow a linked Markdown wiki with cited answers from ingested documents.

- Repository: https://github.com/openapi/gknowledge
- Page: https://hostedmcp.org/servers/openapi/gknowledge
- Transport: stdio
- Stars: 29
- License: MIT
- Latest version: 0.3.0

## Install

### Claude Code

```sh
claude mcp add gknowledge -- uvx gknowledge-mcp
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "gknowledge": {
      "command": "uvx",
      "args": [
        "gknowledge-mcp"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "gknowledge": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "gknowledge-mcp"
      ]
    }
  }
}
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "gknowledge": {
      "command": "uvx",
      "args": [
        "gknowledge-mcp"
      ]
    }
  }
}
```

### Codex (`~/.codex/config.toml`)

```toml
[mcp_servers.gknowledge]
command = "uvx"
args = ["gknowledge-mcp"]
```

### Gemini CLI (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "gknowledge": {
      "command": "uvx",
      "args": [
        "gknowledge-mcp"
      ]
    }
  }
}
```

## FAQ

### Is an API key required to use gknowledge?

No, it works offline by default with a built-in heuristic engine, though you can connect to OpenAI-compatible endpoints if desired.

### Is the wiki compatible with Obsidian?

Yes, it uses plain Markdown with wikilinks and YAML frontmatter, making the folder compatible with Obsidian.