# HavenClaw

> AI execution engine — translating natural language into cross-chain strategy deployment.

***

## What is HavenClaw?

HavenClaw is Haven's primary AI agent interface. It accepts natural language strategy descriptions, parses intent, and generates executable on-chain operations — all within seconds.

It connects directly to the **OpenClaw** agent framework to leverage tool-calling, multi-chain state awareness, and real-time yield data aggregation.

**Try Now:** [**https://haven.financial/demo/**](https://haven.financial/demo/)

{% hint style="info" %}
HavenClaw understands DeFi-native language. You can say *"chase the best stable yield, avoid Curve"* and it will accurately translate that into a set of protocol exclusions and allocation preferences.
{% endhint %}

***

## Key Features

| Feature                | Details                                                              |
| ---------------------- | -------------------------------------------------------------------- |
| Natural language input | Plain English strategy descriptions → machine-executable allocations |
| Cross-chain awareness  | Unified view across Ethereum, Arbitrum, Base, and more               |
| Auto-rebalancing       | Drift-triggered or time-based portfolio rebalancing                  |
| Strategy templates     | Pre-built strategies for common yield profiles                       |
| Execution simulation   | Preview gas costs and expected outcomes before confirming            |

***

## How It Works

### 1. Parse Intent

Your natural language input is processed by the HavenClaw AI model to extract yield targets, risk preferences, and protocol preferences.

### 2. Fetch Live Data

HavenClaw queries live APY data, liquidity depth, and HavenScore risk ratings across all supported protocols.

### 3. Generate Allocation Plan

An optimal allocation is calculated and presented for user confirmation, with simulated outcomes and gas estimates.

### 4. Execute On-chain

Upon confirmation, HavenClaw submits the transaction bundle across chains via the cross-chain router.

***

## Example

```
// Input prompt
"Maximize stablecoin yield with low risk.
 Prefer T-Bill exposure where APY > 4.5%.
 Auto-compound daily, rebalance if drift > 5%."

// Generated strategy
{
  "strategy": "Conservative Fixed Income",
  "allocations": {
    "T-Bills (tokenized)": "55%",
    "USDC lending":        "30%",
    "delta-neutral LP":    "15%"
  },
  "rebalance_trigger": "5% drift",
  "compound_freq":     "daily",
  "estimated_apy":     "6.2–7.8%"
}
```

***

## Related

* [HavenScore — Risk monitoring for every position](/haven-ai-docs/core-products/havenscore.md)
* [OpenClaw Framework — The agent layer powering HavenClaw](/haven-ai-docs/integration/openclaw.md)
* [Quick Start — Deploy your first strategy](/haven-ai-docs/getting-started/quickstart.md)


---

# 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://haven-11.gitbook.io/haven-ai-docs/core-products/havenclaw.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.