Skip to content

Commit

Permalink
Merge pull request #89 from brnaba-aws/main
Browse files Browse the repository at this point in the history 
Updated tool handler and usage doc for python
  • Loading branch information
@brnaba-aws
brnaba-aws authored Nov 20, 2024
2 parents 181534a + 1fdcd36 commit 0cdc4cb
Show file tree
Hide file tree
Showing 9 changed files with 199 additions and 258 deletions.
0 ...ntent/docs/cookbook/examples/api-agent.md → ...tent/docs/cookbook/examples/api-agent.mdx
View file
File renamed without changes.
205 changes: 183 additions & 22 deletions ...ontent/docs/cookbook/tools/weather-api.md → ...ntent/docs/cookbook/tools/weather-api.mdx
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,15 +5,17 @@ description: Understanding Tool use in Bedrock LLM Agent

This guide demonstrates how to create a specialized weather agent using BedrockLLMAgent and a custom weather tool. We'll walk through the process of defining the tool, setting up the agent, and integrating it into your Multi-Agent Orchestrator system.

<br>


1. **Define the Weather Tool**

Let's break down the weather tool definition into its key components:

A. Tool Description
**A. Tool Description**

import { Tabs, TabItem } from '@astrojs/starlight/components';

<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript
export const weatherToolDescription = [
{
Expand All @@ -40,6 +42,35 @@ export const weatherToolDescription = [
}
];
```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
weather_tool_description = [{
"toolSpec": {
"name": "Weather_Tool",
"description": "Get the current weather for a given location, based on its WGS84 coordinates.",
"inputSchema": {
"json": {
"type": "object",
"properties": {
"latitude": {
"type": "string",
"description": "Geographical WGS84 latitude of the location.",
},
"longitude": {
"type": "string",
"description": "Geographical WGS84 longitude of the location.",
},
},
"required": ["latitude", "longitude"],
}
},
}
}]

```
</TabItem>
</Tabs>

**Explanation:**
- This describes the tool's interface to the LLM.
Expand All @@ -49,11 +80,9 @@ export const weatherToolDescription = [
- Requires `latitude` and `longitude` as strings.
- This schema helps the LLM understand how to use the tool correctly.

<br>
<br>

B. Custom Prompt

**B. Custom Prompt**
<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript
export const WEATHER_PROMPT = `
You are a weather assistant that provides current weather data for user-specified locations using only
Expand All @@ -72,6 +101,29 @@ To use the tool, you strictly apply the provided tool specification.
- Complete the entire process until you have all required data before sending the complete response.
`;
```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
weather_tool_prompt = """
You are a weather assistant that provides current weather data for user-specified locations using only
the Weather_Tool, which expects latitude and longitude. Infer the coordinates from the location yourself.
If the user provides coordinates, infer the approximate location and refer to it in your response.
To use the tool, you strictly apply the provided tool specification.
- Explain your step-by-step process, and give brief updates before each step.
- Only use the Weather_Tool for data. Never guess or make up information.
- Repeat the tool use for subsequent requests if necessary.
- If the tool errors, apologize, explain weather is unavailable, and suggest other options.
- Report temperatures in °C (°F) and wind in km/h (mph). Keep weather reports concise. Sparingly use
emojis where appropriate.
- Only respond to weather queries. Remind off-topic users of your purpose.
- Never claim to search online, access external data, or use tools besides Weather_Tool.
- Complete the entire process until you have all required data before sending the complete response.
"""
```
</TabItem>
</Tabs>


**Explanation:**
- This prompt sets the behavior and limitations for the LLM.
Expand All @@ -83,13 +135,11 @@ To use the tool, you strictly apply the provided tool specification.
- Format responses consistently (units, conciseness).
- Stay on topic and use only the provided tool.

<br>
<br>

C. Tool Handler

**C. Tool Handler**
<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript

import { ConversationMessage, ParticipantRole } from "multi-agent-orchestrator";


Expand Down Expand Up @@ -123,20 +173,63 @@ export async function weatherToolHandler(response, conversation: ConversationMes
return message;
}
```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
import requests
from requests.exceptions import RequestException
from typing import List, Dict, Any
from multi_agent_orchestrator.types import ConversationMessage, ParticipantRole

async def weather_tool_handler(response: ConversationMessage, conversation: List[Dict[str, Any]]) -> ConversationMessage:
response_content_blocks = response.content

# Initialize an empty list of tool results
tool_results = []

if not response_content_blocks:
raise ValueError("No content blocks in response")

for content_block in response_content_blocks:
if "text" in content_block:
# Handle text content if needed
pass

if "toolUse" in content_block:
tool_use_block = content_block["toolUse"]
tool_use_name = tool_use_block.get("name")

if tool_use_name == "Weather_Tool":
tool_response = await fetch_weather_data(tool_use_block["input"])
tool_results.append({
"toolResult": {
"toolUseId": tool_use_block["toolUseId"],
"content": [{"json": {"result": tool_response}}],
}
})

# Embed the tool results in a new user message
message = ConversationMessage(
role=ParticipantRole.USER.value,
content=tool_results)

return message
```
</TabItem>
</Tabs>

**Explanation:**
- This handler processes the LLM's request to use the Weather_Tool.
- It iterates through the response content, looking for tool use blocks.
- When it finds a Weather_Tool use:
- It calls `fetchWeatherData` with the provided coordinates.
- It formats the result into a tool result object.
- Finally, it adds the tool results to the conversation as a new user message.

<br>
<br>
- Finally, it returns the tool results to the caller as a new user message.

D. Data Fetching Function
**D. Data Fetching Function**

<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript
async function fetchWeatherData(inputData: { latitude: number; longitude: number }) {
const endpoint = "https://api.open-meteo.com/v1/forecast";
Expand All @@ -158,6 +251,34 @@ async function fetchWeatherData(inputData: { latitude: number; longitude: number
}
}
```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
async def fetch_weather_data(input_data):
"""
Fetches weather data for the given latitude and longitude using the Open-Meteo API.
Returns the weather data or an error message if the request fails.
:param input_data: The input data containing the latitude and longitude.
:return: The weather data or an error message.
"""
endpoint = "https://api.open-meteo.com/v1/forecast"
latitude = input_data.get("latitude")
longitude = input_data.get("longitude", "")
params = {"latitude": latitude, "longitude": longitude, "current_weather": True}

try:
response = requests.get(endpoint, params=params)
weather_data = {"weather_data": response.json()}
response.raise_for_status()
return weather_data
except RequestException as e:
return e.response.json()
except Exception as e:
return {"error": type(e), "message": str(e)}
```
</TabItem>
</Tabs>

**Explanation:**
- This function makes the actual API call to get weather data.
Expand All @@ -174,13 +295,13 @@ These components work together to create a functional weather tool:
4. The fetch function retrieves real weather data based on the LLM's input.

This setup allows the BedrockLLMAgent to provide weather information by seamlessly integrating external data into its responses.
<br>
<br>


2. **Create the Weather Agent**

Now that we have our weather tool defined and the code above in a file called `weatherTool.ts`, let's create a BedrockLLMAgent that uses this tool.

<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript
// weatherAgent.ts

Expand All @@ -206,13 +327,34 @@ const weatherAgent = new BedrockLLMAgent({
});

weatherAgent.setSystemPrompt(WEATHER_PROMPT);

```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
from tools import weather_tool
from multi_agent_orchestrator.agents import (BedrockLLMAgent, BedrockLLMAgentOptions)

weather_agent = BedrockLLMAgent(BedrockLLMAgentOptions(
name="Weather Agent",
streaming=False,
description="Specialized agent for giving weather condition from a city.",
tool_config={
'tool':weather_tool.weather_tool_description,
'toolMaxRecursions': 5,
'useToolHandler': weather_tool.weather_tool_handler
}
))
weather_agent.set_system_prompt(weather_tool.weather_tool_prompt)
```
</TabItem>
</Tabs>

3. **Add the Weather Agent to the Orchestrator**

Now we can add our weather agent to the Multi-Agent Orchestrator:

<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript

import { MultiAgentOrchestrator } from "multi-agent-orchestrator";
Expand All @@ -222,11 +364,23 @@ const orchestrator = new MultiAgentOrchestrator();
orchestrator.addAgent(weatherAgent);

```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
from multi_agent_orchestrator.orchestrator import MultiAgentOrchestrator

orchestrator = MultiAgentOrchestrator()

orchestrator.add_agent(weather_agent)
```
</TabItem>
</Tabs>
## 4. Using the Weather Agent

Now that our weather agent is set up and added to the orchestrator, we can use it to get weather information:

<Tabs syncKey="runtime">
<TabItem label="TypeScript" icon="seti:typescript" color="blue">
```typescript

const response = await orchestrator.routeRequest(
Expand All @@ -235,6 +389,13 @@ const response = await orchestrator.routeRequest(
"session456"
);
```
</TabItem>
<TabItem label="Python" icon="seti:python">
```python
response = await orchestrator.route_request("What's the weather like in New York City?", "user123", "session456")
```
</TabItem>
</Tabs>

### How It Works

Expand Down
Loading

0 comments on commit 0cdc4cb

Please sign in to comment.