Refactor: modularize and streamline LLM providers and utility functions

- Migrate `compress_image.go` to `internal/imageutil` for better encapsulation.
- Reorganize LLM provider implementations into distinct packages (`google`, `openai`, and `anthropic`).
- Replace `go_llm` package name with `llm`.
- Refactor internal APIs for improved clarity, including renaming `anthropic` to `anthropicImpl` and `google` to `googleImpl`.
- Add helper methods and restructure message handling for better separation of concerns.

request.go

@@ -1,17 +1,20 @@
-package go_llm
-
-import (
-	"github.com/openai/openai-go"
-)
-
-type rawAble interface {
-	toRaw() map[string]any
-	fromRaw(raw map[string]any) Input
-}
+package llm
 
+// Input is the interface for conversation inputs.
+// Types that implement this interface can be part of a conversation:
+// Message, ToolCall, ToolCallResponse, and ResponseChoice.
 type Input interface {
-	toChatCompletionMessages(model string) []openai.ChatCompletionMessageParamUnion
+	// isInput is a marker method to ensure only valid types implement this interface.
+	isInput()
 }
+
+// Implement Input interface for all valid input types.
+func (Message) isInput()          {}
+func (ToolCall) isInput()         {}
+func (ToolCallResponse) isInput() {}
+func (ResponseChoice) isInput()   {}
+
+// Request represents a request to a language model.
 type Request struct {
 	Conversation []Input
 	Messages     []Message