custom_provider

README

Custom Provider Example

This example demonstrates how external Go developers can implement their own LLM backend and plug it seamlessly into VibeCoding's advanced Agent framework.

How it Works

VibeCoding provides a public Provider interface:

type Provider interface {
	Chat(ctx context.Context, params ChatParams) <-chan StreamEvent
	Name() string
	Models() []ModelInfo
	GetModel(id string) *ModelInfo
}

In this example, we:

1. Implement a CustomProvider struct that embeds agent.BaseProvider.
2. Define a custom mock model (mock-model).
3. Implement the Chat method to stream mock replies.
4. Integrate keywords to trigger:
   • Text Streaming: Simulates a standard delta-by-delta response stream.
   • Tool Execution Request: Simulates the LLM deciding to read a file (README.md) by sending a StreamToolCall event with a ToolCallBlock.
5. Run the VibeCoding agent loop with this custom provider.

Notice how VibeCoding's built-in tool registry automatically intercepts the LLM's StreamToolCall event, runs the actual read tool against the workspace safely, and provides the result back to the assistant!

How to Run

1. Make sure you are in the root directory of the project.
2. Run the example:

   go run example/custom_provider/main.go