# LingNao

> LingNao is a free MCP server for ai models and llm tooling. Deterministic reasoning gateway that recomputes numbers and logic, abstaining when unsure. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Free
- **Platforms:** self-hosted, Web, API, CLI
- **Website:** https://hongchenlingjing.com/verify/
- **Repository:** https://github.com/genesis-plan/lingnao
- **MCP Registry name:** io.github.genesis-plan/lingnao
- **Categories:** AI models and LLM tooling, developer tools
- **Listing updated:** 2026-11-15

## About

One gateway tool, `lingnao`, takes an `op` name and arguments and returns structured JSON. The README lists 51 operations covering A* and deterministic MCTS planning, causal reasoning with do-calculus, PAC sample bounds, world models and a set of mathematical trap tests built on classic theorems. The same input always gives the same output.

Answers have three states: true, false, or unverifiable, where the engine abstains rather than guessing. It is not a language model and generates no text. An optional LLM can be used only to turn natural language into JSON or to explain results, and its output is labelled UNVERIFIED_LLM and kept out of the reasoning chain.

The operator also runs a web checker where you paste a conclusion containing numbers or formulas and get right, wrong or unverifiable back, with a report id that can be recomputed offline. It deliberately stays out of medical, legal, financial and similar advice. A hosted MCP endpoint is listed at ¥0.01 per call and is currently open for free. The local server runs from npm with `npx -y lingnao-mcp`, and much of the documentation is in Chinese.

## Features

- **Single gateway tool:** One tool dispatches about 51 operations by name.
- **Three-state verdicts:** Returns true, false or unverifiable and abstains when it cannot decide.
- **Deterministic results:** Identical input and kernel version give identical output.
- **Planning and causality:** Includes A* and MCTS planning and do-calculus causal inference.
- **Offline-checkable reports:** Report ids are a SHA-256 of input and kernel version.
- **Hosted or local:** Runs locally from npm or through a hosted endpoint.

## Good for

- Recompute arithmetic and formulas in an agent's answer
- Check equations against a claimed value
- Plan tasks with a deterministic search
- Run causal queries on a small graph
- Keep a reproducible reasoning trail for an agent decision

## Install

### Claude Code (Terminal)

```sh
claude mcp add lingnao -- npx -y lingnao-mcp
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "lingnao": {
      "command": "npx",
      "args": [
        "-y",
        "lingnao-mcp"
      ]
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "lingnao": {
      "command": "npx",
      "args": [
        "-y",
        "lingnao-mcp"
      ]
    }
  }
}
```

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

```json
{
  "servers": {
    "lingnao": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "lingnao-mcp"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "lingnao": {
      "command": "npx",
      "args": [
        "-y",
        "lingnao-mcp"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "lingnao": {
      "command": "npx",
      "args": [
        "-y",
        "lingnao-mcp"
      ]
    }
  }
}
```

## Tools

- `lingnao`: Single gateway tool; the op argument selects one of about 51 operations

## FAQ

### Does it use a language model?

The core does not. An optional LLM may translate natural language into JSON or explain results, and its output is marked unverified.

### What does it refuse to check?

The web checker lists medical, mental-health, education, investment, financial and legal advice as outside its scope.

### How do I run it?

Use npx -y lingnao-mcp as the server command in your MCP client.

### Is there a hosted version?

Yes. The hosted MCP endpoint is listed at ¥0.01 per call and is currently open for free.

### Is LingNao free?

Yes. LingNao is free to use.

### How do I add LingNao to Claude Code?

Run this command in your terminal: claude mcp add lingnao -- npx -y lingnao-mcp

### What are the best alternatives to LingNao?

The closest alternatives to LingNao are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/lingnao
