# System Architecture

**How Clawd Bot Works**

Clawd Bot functions as a bridge between natural language interfaces and your local operating system. The architecture is modular, allowing for flexibility in how the brain, the body, and the interface interact.

**The Interface Layer**\
This is where you communicate with the agent. Clawd Bot supports multiple providers simultaneously. You can send a message via Telegram from your phone, and the agent receives it on your desktop at home. The interface layer normalizes these inputs into a standard format that the core logic can understand.

**The Agent Core**\
The brain of the operation. The core processes your intent using a Large Language Model. You can configure this to use external providers like Anthropic and OpenAI, or keep it entirely offline with local models like Llama or Mistral. The core analyzes the request and determines the necessary steps to fulfill it.

**The Tool Belt**\
Once the core formulates a plan, it utilizes the Tool Belt. This is a collection of capabilities granted to the agent.

* **File System Tool** enables reading, writing, and organizing local files.
* **Browser Tool** launches a headless web browser to interact with websites.
* **Shell Tool** executes command line instructions for system management.

**The Execution Loop**\
Clawd Bot operates on a continuous feedback loop known as Think Plan Act Observe.

1. **Think** The agent analyzes the user request.
2. **Plan** It breaks the request down into a series of actionable steps.
3. **Act** It performs the first step using the appropriate tool.
4. **Observe** It reads the output of that action to verify success or identify errors.


---

# 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.clawdtoken.xyz/1-overview/system-architecture.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.