# Prompts for Onchain MCP Server
This directory contains prompt templates for various blockchain analysis tasks.
## Structure
```
prompts/
├── __init__.py
├── contract_audit.py # Contract analysis prompts
├── token_analysis.py # Token research prompts (future)
├── defi_analysis.py # DeFi protocol prompts (future)
└── README.md # This file
```
## Usage
### Contract Audit Prompts
The contract audit tool now supports multiple output formats:
#### 1. Raw Data (Default)
```python
result = contract_audit_tool(address, network, format="raw")
# Returns: Raw JSON data
```
#### 2. Security Audit Prompt
```python
result = contract_audit_tool(address, network, format="audit")
# Returns: Structured security audit prompt
```
#### 3. Quick Analysis Prompt
```python
result = contract_audit_tool(address, network, format="quick")
# Returns: Quick assessment prompt
```
#### 4. Deep Dive Prompt
```python
result = contract_audit_tool(address, network, format="deep")
# Returns: Comprehensive analysis prompt
```
## Prompt Response Format
When using prompt formats, the response includes:
```json
{
"prompt": "Formatted prompt template with data filled in",
"context": { /* Raw audit data */ },
"format_type": "audit|quick|deep",
"address": "0x...",
"network": "base",
"timestamp": "2024-01-01T00:00:00Z"
}
```
## Adding New Prompts
1. Create a new prompt file in this directory
2. Define prompt functions that return formatted strings
3. Import and register in `server.py`
4. Update tools to use the new prompts
Example:
```python
# prompts/token_analysis.py
def token_research_prompt():
return "Research this token: {address} on {network}..."
# server.py
from prompts.token_analysis import token_research_prompt
mcp.prompt()(token_research_prompt)
```
