llm

type AssistantResult struct {
	Content    string
	ToolCalls  []llms.ToolCall
	ModelName  string
	StopReason string
	Usage      *Usage
}

type ChatTemplate interface {
	Format(ctx context.Context, data map[string]any) ([]llms.MessageContent, error)
}

type FakeModel struct {
	// contains filtered or unexported fields
}

type ResponseScript struct {
	Content string
	// Chunks, when non-empty, are emitted one-by-one via StreamingFunc instead of Content as a single chunk.
	Chunks []string
	// ReasoningChunks are emitted via StreamingReasoningFunc when configured.
	ReasoningChunks []string
	ToolCalls       []llms.ToolCall
	Err             error
	// ErrAfterStream is returned after streaming deltas have been delivered.
	ErrAfterStream error
}

type RetryConfig struct {
	// MaxAttempts is the total number of execution attempts. Zero uses DefaultRetryConfig.
	MaxAttempts int
	// InitialInterval is the delay before the first retry. Zero uses DefaultRetryConfig.
	InitialInterval time.Duration
	// MaxInterval caps the delay between retries. Zero uses DefaultRetryConfig.
	MaxInterval time.Duration
	// Multiplier controls delay growth. Zero uses DefaultRetryConfig.
	Multiplier float64
	// OnRetry is called before each retry attempt. May be nil.
	OnRetry func(attempt int, delay time.Duration, err error)
}

type StreamOptions struct {
	ModelName        string
	Temperature      float64
	MaxTokens        int
	Tools            []llms.Tool
	OnTextDelta      func(delta string) error
	OnReasoningDelta func(delta string) error
	// Retry overrides the default transient retry policy when non-zero MaxAttempts is set.
	Retry RetryConfig
}

type Usage struct {
	PromptTokens     int
	CompletionTokens int
	TotalTokens      int
	CacheRead        int
	CacheWrite       int
}