veryfront 0.1.283 → 0.1.284

package/src/src/extensions/interfaces/ai-model-provider.ts

@@ -1,100 +0,0 @@
-/**
- * Contract interface for AI/LLM model providers.
- *
- * Default implementation: `@veryfront/ext-openai`
- *
- * @module extensions/interfaces/ai-model-provider
- */
-
-/** A single part of a multi-modal message (text or image). */
-export interface ContentPart {
-  /** Part type. */
-  type: "text" | "image";
-  /** Text content (when `type` is `"text"`). */
-  text?: string;
-  /** Image URL or base64 data (when `type` is `"image"`). */
-  imageUrl?: string;
-}
-
-/** A chat message in a conversation. */
-export interface ChatMessage {
-  /** Message role. */
-  role: "system" | "user" | "assistant" | "tool";
-  /** Simple text content. */
-  content?: string;
-  /** Multi-modal content parts (used instead of `content`). */
-  parts?: ContentPart[];
-  /** Tool call ID this message responds to (for `tool` role). */
-  toolCallId?: string;
-}
-
-/** Definition of a tool the model may invoke. */
-export interface ToolDefinition {
-  /** Tool name. */
-  name: string;
-  /** Human-readable description of what the tool does. */
-  description: string;
-  /** JSON Schema describing the tool's parameters. */
-  parameters: Record<string, unknown>;
-}
-
-/** Options passed to {@link AIModelProvider.complete} and {@link AIModelProvider.stream}. */
-export interface CompletionOptions {
-  /** Model identifier (e.g. `"gpt-4o"`, `"claude-sonnet-4-20250514"`). */
-  model: string;
-  /** Conversation messages. */
-  messages: ChatMessage[];
-  /** Sampling temperature. */
-  temperature?: number;
-  /** Maximum tokens to generate. */
-  maxTokens?: number;
-  /** Available tools for function calling. */
-  tools?: ToolDefinition[];
-  /** Additional provider-specific options. */
-  [key: string]: unknown;
-}
-
-/** Result returned from {@link AIModelProvider.complete}. */
-export interface CompletionResult {
-  /** Generated text content. */
-  content: string;
-  /** Tool calls requested by the model, if any. */
-  toolCalls?: Array<{
-    id: string;
-    name: string;
-    arguments: string;
-  }>;
-  /** Token usage statistics. */
-  usage?: {
-    promptTokens: number;
-    completionTokens: number;
-    totalTokens: number;
-  };
-}
-
-/** A chunk emitted during streaming completion. */
-export interface StreamChunk {
-  /** Incremental text delta. */
-  content?: string;
-  /** Incremental tool call delta. */
-  toolCallDelta?: {
-    id?: string;
-    name?: string;
-    arguments?: string;
-  };
-  /** Whether this is the final chunk. */
-  done: boolean;
-}
-
-/**
- * AIModelProvider contract interface.
- *
- * Implementations provide chat completion and streaming capabilities
- * against large language models.
- */
-export interface AIModelProvider {
-  /** Generate a complete response for the given conversation. */
-  complete(options: CompletionOptions): Promise<CompletionResult>;
-  /** Stream a response token-by-token. */
-  stream(options: CompletionOptions): AsyncIterable<StreamChunk>;
-}