# DragonClaw

DragonClaw is a self-hosted, open-source AI agent built for crypto-native Chinese users.

It ships with Binance trading skills from day one. It also connects to WeChat, DingTalk, Feishu, Telegram, Discord, and custom clients.

Use Chinese LLMs like DeepSeek, Qwen, Kimi, and GLM at a fraction of the cost of many Western models.

[Official website](https://www.dragonclaw.asia)

### Start here

* [Quick Start](https://docs.dragonclaw.asia/getting-started/quickstart)
* [Built-in Skills Overview](https://docs.dragonclaw.asia/built-in-skills/overview)
* [REST API](https://docs.dragonclaw.asia/api-reference/rest-api)
* [Security Model](https://docs.dragonclaw.asia/security/security-model)

### Why teams choose DragonClaw

* **Crypto native**. Seven Binance skills ship built in.
* **Chinese ecosystem ready**. Native workflows for Chinese chat platforms.
* **Self-hosted**. Your data, keys, and memory stay on your infrastructure.
* **OpenClaw compatible**. Existing `SKILL.md` skills keep working.
* **Production hardened**. Auth, retries, rate limits, logging, and health checks are built in.

### Quick start

{% tabs %}
{% tab title="Install script" %}

```bash
curl -fsSL https://dragonclaw.asia/install.sh | bash
dragonclaw onboard
dragonclaw start
```

{% endtab %}

{% tab title="npm" %}

```bash
npm i -g dragonclaw
dragonclaw onboard
dragonclaw start
```

{% endtab %}
{% endtabs %}

### How it works

DragonClaw runs as a layered runtime.

1. **Client channels** receive messages from WeChat, DingTalk, Feishu, Telegram, Discord, WebSocket clients, or the CLI.
2. **Gateway layer** accepts REST and WebSocket traffic, then applies auth, rate limits, and input validation.
3. **Agent layer** selects relevant skills, retrieves memory, calls the model, and handles tool execution.
4. **Execution layer** combines skills, LLM providers, memory, and connectors to produce the final response.

### Core capabilities

* Real-time meme token tracking
* Smart money ranking and wallet analysis
* Token metadata and contract risk checks
* Binance spot trading with confirmation gates
* Persistent memory across channels and sessions

{% hint style="info" %}
Start with **Quick Start** if you want the fastest path to a working agent.
{% endhint %}


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.dragonclaw.asia/readme.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.