# douglasgan/asktian [Health: Active]

**Category:** 🔮 Spirituality & Esoterica  
**Repository:** https://github.com/douglasgan/asktian-mcp  
**GitHub Stars:** 2  
**npm Downloads (last month):** 294  
**Views:** 4  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/douglasgan-asktian

## Description
Chinese metaphysics (bazi 八字, qimen 奇門, 5-element, daily 干支) as decision-support tools. Ask "when should I do X" and get specific time windows instead of vague advice. 5 tools: daily reading, compat, best-time-for-action, today's energy, name analysis. npm install -g @asktian/mcp-server

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "asktian": {
    "command": "npx",
    "args": ["-y","@asktian/mcp-server"]
  }
}
```

## Documentation

## What douglasgan/asktian MCP server does

douglasgan/asktian MCP server exposes six MCP tools based on Chinese metaphysics concepts, including bazi, qimen, the five elements, and daily 干支 energy. It is designed for AI assistants that need to answer questions about personal daily guidance, relationships, names, action timing, or a binary prediction-market question.

The server can provide a personalized daily reading from a birthdate, optional birth time, and gender. It can compare two people across love, career, friendship, or general compatibility. It can also identify three recommended time windows for an action over a configurable number of days, with reasons and periods to avoid. Other tools describe the general energy of a date, analyze a name, or return a deterministic metaphysical signal for a yes/no market question.

## How it works

The douglasgan/asktian MCP server communicates with clients through MCP over JSON-RPC. In its default form, an MCP client launches the npm-provided command as a subprocess and sends tool calls over stdio. A streamable-HTTP mode is also available for agents that connect to a network endpoint.

Most reading tools call the live askTIAN backend. They work without a key, subject to an approximate limit of 300 requests per 15 minutes per IP. An optional API key increases limits and enables the premium `tian.*` reading tier. The market-read tool is computed locally as a deterministic novelty signal, while the best-time tool currently returns a guided preview rather than using a backend endpoint.

## Setup and configuration

Install the package globally with `npm install -g @asktian/mcp-server`, then register `asktian-mcp` in the MCP configuration for the chosen client. The README provides configurations for Claude Desktop, Cursor, Windsurf, and Zed. Other MCP clients can launch the same command as a subprocess.

The `ASKTIAN_API_KEY` variable is optional. Set it in the client environment when higher request limits or the premium reading tier are needed; omit it for the default no-key mode. For HTTP operation, set `ASKTIAN_MCP_HTTP=1` and optionally choose a `PORT`; the example uses port 8080. The hosted endpoint is `https://mcp.asktian.com/mcp`, where the key is supplied as a Bearer authorization header.

## Tools and capabilities

The douglasgan/asktian MCP server provides these tools:

- `asktian_daily_reading`: Returns a personal daily reading with an archetype, stem-branch energy, favorable colors, direction, hours, and written guidance.
- `asktian_compatibility`: Compares two birthdates and returns a qualitative relationship label, category, element flow, advice, and conditional score information.
- `asktian_best_time_for_action`: Finds up to three recommended date-and-hour windows for an action and identifies windows to avoid.
- `asktian_today_energy`: Describes a date's stem and branch plus its dominant five-element character; the date defaults to today.
- `asktian_name_analysis`: Produces a short energetic profile for an English or Chinese name, with automatic language detection available.
- `asktian_market_read`: Returns a yes, no, or neutral signal, score, confidence, reasoning, and disclaimer for a binary prediction-market question.

## Limitations and notes

These outputs are metaphysical decision-support readings, not objective forecasts. The market tool is explicitly presented as entertainment and not financial advice. Its signal is deterministic and intended as a novelty that can be evaluated over time.

Compatibility scores below 60 may be hidden, with the qualitative label taking precedence. Clients should respect the response's `should_show_score` and `note_for_ai` guidance rather than exposing a suppressed number. The API is rate-limited without a key, and the best-time feature is currently a guided preview. LangChain and LlamaIndex support the server through their native MCP adapters, with setup examples supplied in the repository documentation.

_Full upstream README: https://allmcps.com/mcp/douglasgan-asktian/readme_