Examples

This section provides practical examples to help you understand how to use Warden and its features effectively. From running agents locally to exporting and importing agents, these examples cover a wide range of use cases.

Example 1: Running GPT-4-Free Locally

Step 1: Start the GPT-4-Free API

Ensure that the GPT-4-Free API is running locally. Start it using:

g4f api

Step 2: Run the Warden Framework

Launch the Warden application:

go run main.go

Step 3: Select CLI Chatbot

From the interactive menu, select the CLI Chatbot option and start chatting with the agent powered by GPT-4-Free.

Example 2: Exporting and Importing Agents

Exporting an Agent

Agents can be serialized into .agent files for reuse or sharing. For example:

agent.ExportToFile("my_agent.agent")

This command saves the agent’s configuration and memory to a file named my_agent.agent.

Importing an Agent

To import an agent from a .agent file:

loadedAgent := agent.ImportFromFile("my_agent.agent")

The imported agent is now ready for use.

Example 3: Using a Custom Tool

Step 1: Create a Custom Tool

Define a custom tool for retrieving weather data:

type WeatherTool struct {}

func (w *WeatherTool) Execute(input string) string {
    return "The weather today is sunny."
}

Step 2: Register the Tool

Register the tool with the agent:

agent.RegisterTool("weather", &WeatherTool{})

Step 3: Use the Tool

Invoke the tool during an agent interaction:

response := agent.UseTool("weather", "What’s the weather?")
fmt.Println(response) // Outputs: The weather today is sunny.

Example 4: Multi-Agent Collaboration

Step 1: Create Two Agents

Create two agents with distinct capabilities:

agent1 := NewAgent("GPT-4 Agent")
agent2 := NewAgent("Claude Agent")

Step 2: Share Memory Between Agents

Enable shared memory so agents can collaborate:

sharedMemory := NewSharedMemory()
agent1.SetMemory(sharedMemory)
agent2.SetMemory(sharedMemory)

Step 3: Collaborate on a Task

Agent 1 stores data:

agent1.Memory.Save("task", "Complete project report")

Agent 2 retrieves and uses the data:

task := agent2.Memory.Retrieve("task")
fmt.Println(task) // Outputs: Complete project report

Example 5: Posting Tweets via Twitter Integration

Step 1: Configure Twitter API Credentials

Set your Twitter API credentials as environment variables:

export TWITTER_API_KEY=your_twitter_api_key
export TWITTER_API_SECRET=your_twitter_api_secret

Step 2: Use the Twitter Tool

Generate and post a tweet:

tweet := agent.GenerateResponse("What should I tweet today?")
twitterTool.PostTweet(tweet)

Conclusion

These examples demonstrate the flexibility and power of Warden for building AI-powered agents. Experiment with these use cases and adapt them to suit your specific requirements.