Add AI/LLM integration guide with prompt engineering examples

Problem

While the library is designed for AI agents, there's no guide on how to effectively integrate with LLM frameworks or write prompts for safe agent behavior.

Current State

• Tool definitions exist but lack context for LLM usage
• No examples of system prompts for safe transaction handling
• No guidance on handling LLM hallucinations in financial context

Recommendation

Add documentation covering:

1. System Prompt Templates

   • Safe transaction handling guidelines
   • How to interpret tool responses
   • When to ask for human confirmation

2. Framework Integration Examples

   • OpenAI function calling
   • Anthropic tool use
   • LangChain/LlamaIndex integration

3. Safety Guidelines for LLMs

   • Validating amounts before sending
   • Double-checking addresses
   • Handling ambiguous user requests

4. Example Prompts

## System Prompt Example

You are a financial assistant with access to an Ethereum wallet.

SAFETY RULES:
- Always confirm the recipient address with the user before sending
- For amounts > $100, summarize the transaction and ask for confirmation
- Never send to addresses the user hasn't explicitly provided
- If unsure about any detail, ask for clarification

Priority

Medium - Important for target use case (AI agents)

Start by reviewing the existing tool definitions and the repository's documentation entry points to understand the current usage context. Cover the requested system prompt templates, OpenAI and Anthropic tool usage, LangChain/LlamaIndex examples, and financial safety guidance. Done means the guide includes practical prompts and integration examples for each listed area.