proxy

type APIKeyRoutingConfig struct {
	AllowedProviders   []string
	AllowedModels      []string
	ModelOverrides     map[string]string
	RateLimitTier      string
	CredentialStrategy byok.Strategy
}

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

type CacheConfig struct {
	// Enable caching for different endpoints
	EnableChatCache      bool `env:"ENABLE_CHAT_CACHE,default=true"`
	EnableEmbeddingCache bool `env:"ENABLE_EMBEDDING_CACHE,default=true"`
	EnableModelCache     bool `env:"ENABLE_MODEL_CACHE,default=true"`
	EnableProviderCache  bool `env:"ENABLE_PROVIDER_CACHE,default=true"`
	// Skip cache for specific models or patterns
	SkipCacheModels []string `env:"SKIP_CACHE_MODELS"`
	// Force cache refresh header
	CacheControlHeader string `env:"CACHE_CONTROL_HEADER,default=X-Cache-Control"`
}

type CacheCost struct {
	WriteTokens float64 `json:"write_tokens"` // Cost of writing tokens to cache
	ReadTokens  float64 `json:"read_tokens"`  // Cost of reading tokens from cache
	TotalCost   float64 `json:"total_cost"`   // Total cost including cache operations
}

type CacheManager interface {
	GetModel(ctx context.Context, key string) (any, bool, error)
	SetModel(ctx context.Context, key string, value any) error
	GetResponse(ctx context.Context, key string) ([]byte, bool, error)
	SetResponse(ctx context.Context, key string, response []byte) error
}

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

type CacheStatusProvider interface {
	GetCacheStatus() string
	GetCacheAge() int // Returns cache age in seconds, or 0 if not cached
}

type ChatCompletionRequest struct {
	Request  inference.ChatRequest
	Route    string
	Provider *ProviderPreferences

	// Internal fields
	APIKey       string               `json:"-"`
	TenantID     string               `json:"-"`
	APIKeyConfig *APIKeyRoutingConfig `json:"-"`
	RequestID    string               `json:"-"`
}

type ChatCompletionResponse struct {
	Response inference.ChatResponse

	// Internal fields (not serialized)
	CacheStatus string     `json:"-"`
	CacheAge    int        `json:"-"` // Seconds since cached
	ETag        string     `json:"-"` // Entity tag for response
	CacheCost   *CacheCost `json:"-"` // Cache pricing information
}

type ChatCompletionStreamResponse interface {
	// Read returns the next canonical event or io.EOF when done.
	Read() (*inference.StreamEvent, error)

	// Close releases any resources
	Close() error
}

type Config struct {
	// Registry provides access to LLM provider connectors
	Registry *registry.Registry

	// Router handles intelligent model selection and failover
	Router router.ModelRouter

	// CacheManager handles response caching (optional)
	CacheManager *cache.Manager

	// CacheConfig configures caching behavior (optional)
	CacheConfig *CacheConfig

	// Middlewares to apply to the proxy service
	Middlewares []Middleware
}

type EmbeddingsRequest struct {
	Request inference.EmbeddingRequest

	// Internal fields
	APIKey       string               `json:"-"`
	TenantID     string               `json:"-"`
	APIKeyConfig *APIKeyRoutingConfig `json:"-"`
	RequestID    string               `json:"-"`
}

type EmbeddingsResponse struct {
	Response inference.EmbeddingResponse

	// Internal fields (not serialized)
	CacheStatus string     `json:"-"`
	CacheAge    int        `json:"-"` // Seconds since cached
	ETag        string     `json:"-"` // Entity tag for response
	CacheCost   *CacheCost `json:"-"` // Cache pricing information
}

type EndpointInfo struct {
	Provider   string `json:"provider"`
	Endpoint   string `json:"endpoint"`
	Available  bool   `json:"available"`
	Latency    *int   `json:"latency_ms,omitempty"`
	CostPrompt string `json:"cost_prompt,omitempty"`
	CostOutput string `json:"cost_output,omitempty"`
}

type Middleware interface {
	// Wrap wraps the given proxy with the middleware functionality
	Wrap(Proxy) Proxy
}

type MiddlewareFunc func(Proxy) Proxy

type ModelArchitecture struct {
	InputModalities  []string `json:"input_modalities"`
	OutputModalities []string `json:"output_modalities"`
	Tokenizer        string   `json:"tokenizer"`
	InstructType     *string  `json:"instruct_type"`
}

type ModelEndpointsResponse struct {
	Model     string         `json:"model"`
	Endpoints []EndpointInfo `json:"endpoints"`
}

type ModelInfo struct {
	ID            string `json:"id"`
	CanonicalSlug string `json:"canonical_slug,omitempty"`
	Name          string `json:"name,omitempty"`
	Object        string `json:"object"`
	Created       int64  `json:"created"`
	OwnedBy       string `json:"owned_by"`

	// Extended metadata for OpenRouter compatibility
	Pricing             *ModelPricing      `json:"pricing,omitempty"`
	Context             *int               `json:"context_length,omitempty"`
	Type                string             `json:"type,omitempty"`
	Description         string             `json:"description,omitempty"`
	Architecture        *ModelArchitecture `json:"architecture,omitempty"`
	TopProvider         *TopProviderInfo   `json:"top_provider,omitempty"`
	SupportedParameters []string           `json:"supported_parameters,omitempty"`
}

type ModelPricing struct {
	Prompt     string `json:"prompt"`
	Completion string `json:"completion"`
	Currency   string `json:"currency"`
}

type ModelsResponse struct {
	Object string      `json:"object"`
	Data   []ModelInfo `json:"data"`

	// Internal fields (not serialized)
	CacheStatus string `json:"-"`
}

type Option func(*Config)

type Options struct {
	// RequestTimeout is the timeout for individual requests
	RequestTimeout time.Duration

	// EnableMetrics enables metrics collection
	EnableMetrics bool

	// EnableLogging enables request/response logging
	EnableLogging bool
}

type ProviderError struct {
	Provider string
	Code     string
	Message  string
	Err      error
}

type ProviderInfo struct {
	ID              string   `json:"id"`
	Name            string   `json:"name"`
	Description     string   `json:"description,omitempty"`
	URL             string   `json:"url,omitempty"`
	Models          []string `json:"models"`
	Capabilities    []string `json:"capabilities,omitempty"`
	RequiresAuth    bool     `json:"requires_auth"`
	AuthDescription string   `json:"auth_description,omitempty"`
}

type ProviderPreferences struct {
	Order         []string `json:"order,omitempty"`
	Ignore        []string `json:"ignore,omitempty"`
	Only          []string `json:"only,omitempty"`
	AllowFallback bool     `json:"allow_fallback,omitempty"`
}

type ProvidersResponse struct {
	Providers []ProviderInfo `json:"providers"`

	// Internal fields (not serialized)
	CacheStatus string `json:"-"`
}

type Proxy interface {
	// ProcessChatCompletion handles chat completion requests with routing and processing
	ProcessChatCompletion(ctx context.Context, req *ChatCompletionRequest) (*ChatCompletionResponse, error)

	// ProcessChatCompletionStream handles streaming chat completion requests
	ProcessChatCompletionStream(ctx context.Context, req *ChatCompletionRequest) (ChatCompletionStreamResponse, error)

	// ProcessEmbeddings handles embedding generation requests
	ProcessEmbeddings(ctx context.Context, req *EmbeddingsRequest) (*EmbeddingsResponse, error)

	// ListModels returns available models based on routing configuration
	ListModels(ctx context.Context) (*ModelsResponse, error)

	// ListProviders returns available provider information
	ListProviders(ctx context.Context) (*ProvidersResponse, error)

	// GetModelEndpoints returns provider endpoints for a specific model
	GetModelEndpoints(ctx context.Context, modelID string) (*ModelEndpointsResponse, error)
}

type RoutingConfig struct {
	// EnableFailover enables automatic failover to other providers
	EnableFailover bool

	// PreferredProviders is an ordered list of preferred providers
	PreferredProviders []string

	// ExcludedProviders is a list of providers to exclude
	ExcludedProviders []string

	// EnableLoadBalancing enables load balancing across providers
	EnableLoadBalancing bool

	// EnableStickyRouting enables sticky routing for conversations
	EnableStickyRouting bool
}

type RoutingError struct {
	Model  string
	Reason string
	Err    error
}

type SecurityConfig struct {
	// EnableRateLimiting enables rate limiting
	EnableRateLimiting bool

	// RateLimitPerMinute is the number of requests allowed per minute
	RateLimitPerMinute int

	// EnableContentFiltering enables content filtering
	EnableContentFiltering bool

	// BlockedPatterns is a list of regex patterns to block
	BlockedPatterns []string

	// EnableAPIKeyValidation enables API key validation
	EnableAPIKeyValidation bool
}

type TopProviderInfo struct {
	ContextLength       int `json:"context_length"`
	MaxCompletionTokens int `json:"max_completion_tokens"`
}

type ValidationConfig struct {
	// StrictMode enables strict validation of requests
	StrictMode bool

	// MaxTokensLimit is the maximum allowed max_tokens value
	MaxTokensLimit int

	// MaxMessagesLimit is the maximum number of messages allowed
	MaxMessagesLimit int

	// MaxMessageLength is the maximum length of a single message
	MaxMessageLength int
}

type ValidationError struct {
	Field   string
	Message string
}