# gknowledge

Gives your agent a wiki of linked Markdown pages: feed it documents, the model keeps pages current and answers with citations.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 24
- Author: openapi
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/openapi/gknowledge
- Updated: 2026-10-20
- HTML: https://affpapa.org/en/nevibecode/mcp/gknowledge

## What it does

The server exposes a knowledge base shaped like a wiki. Dropped-in documents become linked Markdown pages with YAML frontmatter, and a plain SQLite index sits alongside. You can open the folder in Obsidian or any editor; the files stay on disk.

Lookups combine lexical search with graph expansion across page links, and each answer points back to the pages it used. A generated page keeps track of where its material came from. Whatever the model changes lands in a review queue first, so edits do not stick on their own.

It starts as `uvx gknowledge-mcp`, so uv has to be on the machine. With no model attached, a built-in heuristic engine handles ingestion and answers offline; adding a model is worth it when you want tighter synthesis.

## Requirements
- Needs: Python and uv (commands run via uvx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add gknowledge -- uvx gknowledge-mcp
```
uvx runs the command named after the package; if the README names another one, use uvx --from gknowledge-mcp <command>

### Codex

**Platform CLI**
```bash
codex mcp add gknowledge -- uvx gknowledge-mcp
```
uvx runs the command named after the package; if the README names another one, use uvx --from gknowledge-mcp <command>

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "gknowledge": {
            "command": "uvx",
            "args": [
                "gknowledge-mcp"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from gknowledge-mcp <command>

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "gknowledge": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "gknowledge-mcp"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from gknowledge-mcp <command>

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add gknowledge uvx gknowledge-mcp
```
uvx runs the command named after the package; if the README names another one, use uvx --from gknowledge-mcp <command>

## When to use it
- turn a folder of documents into a wiki that stays current
- answer questions from your own notes with page citations
- review model edits before they reach the knowledge base
- read accumulated knowledge in Obsidian with no export step
- search the base when an outside API is unavailable

## Example prompts
- Ingest the files in `papers` into the wiki and list which pages changed
- Find everything about incremental indexing and answer with citations
- Show the review queue and the pages the model touched last

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33918
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81115
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119160
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47392
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43872

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md