@imenam/unified-llm-client 1.2.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 imenam
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -0,0 +1,233 @@
+# @imenam/unified-llm-client
+
+A unified interface for multiple LLM providers — OpenAI, Gemini, Anthropic, and Qwen — with built-in support for tool-calling, streaming, multimodal inputs, and audio generation.
+
+## Features
+
+- **Unified API** — one interface, four providers
+- **Tool-calling** — automatic execution loop with structured results
+- **Streaming** — chunk-by-chunk responses via callbacks
+- **Multimodal** — image and audio file inputs
+- **Audio generation** — Text-to-Speech (Gemini)
+- **Usage tracking** — token counts and estimated costs
+- **Full TypeScript support** — exported types and `.d.ts` declarations
+
+## Installation
+
+```bash
+npm install @imenam/unified-llm-client
+```
+
+## Supported Providers
+
+| Provider  | Chat | Streaming | Tool-calling | Multimodal | Audio TTS |
+|-----------|:----:|:---------:|:------------:|:----------:|:---------:|
+| OpenAI    | ✅   | ✅        | ✅           | ✅         | ❌        |
+| Gemini    | ✅   | ✅        | ✅           | ✅         | ✅        |
+| Anthropic | ✅   | ✅        | ✅           | ✅         | ❌        |
+| Qwen      | ✅   | ✅        | ✅           | ✅         | ❌        |
+
+## Quick Start
+
+Set your API keys in a `.env` file:
+
+```env
+OPENAI_API_KEY=sk-...
+GOOGLE_API_KEY=...
+ANTHROPIC_API_KEY=sk-ant-...
+DASHSCOPE_API_KEY=...   # for Qwen
+```
+
+```typescript
+import { UnifiedLLMClient, MODELS } from "@imenam/unified-llm-client";
+
+const client = new UnifiedLLMClient(MODELS.GPT_4O_MINI);
+
+const response = await client.chat({
+  message: "What is the capital of France?",
+});
+
+console.log(response.content); // "The capital of France is Paris."
+```
+
+## Available Models
+
+```typescript
+import { MODELS } from "@imenam/unified-llm-client";
+
+// OpenAI
+MODELS.GPT_4O_MINI
+MODELS.GPT_5
+
+// Gemini
+MODELS.GEMINI_FLASH_2_5
+MODELS.GEMINI_2_5_PRO
+MODELS.GEMINI_3_FLASH
+MODELS.GEMINI_3_PRO
+MODELS.GEMINI_3_1_PRO
+
+// Anthropic
+MODELS.CLAUDE_SONNET_4_6
+MODELS.CLAUDE_OPUS_4_6
+
+// Qwen
+MODELS.QWEN_VL_3
+MODELS.QWEN_VL_FLASH
+```
+
+You can also pass a custom model:
+
+```typescript
+const client = new UnifiedLLMClient({
+  provider: "openai",
+  modelName: "gpt-4o",
+});
+```
+
+## Tool-calling
+
+Define tools with an `executor` function — the client will call them automatically and loop until the LLM produces a final text response.
+
+```typescript
+import { UnifiedLLMClient, MODELS } from "@imenam/unified-llm-client";
+import type { ToolDefinition } from "@imenam/unified-llm-client";
+
+const tools: ToolDefinition[] = [
+  {
+    name: "get_weather",
+    description: "Get the current weather for a city",
+    parameters: {
+      type: "object",
+      properties: {
+        city: { type: "string", description: "City name" },
+      },
+      required: ["city"],
+    },
+    executor: async ({ city }) => {
+      return `The weather in ${city} is sunny and 22°C.`;
+    },
+  },
+];
+
+const client = new UnifiedLLMClient(MODELS.GEMINI_FLASH_2_5);
+
+const response = await client.chat({
+  message: "What's the weather like in Paris?",
+  tools,
+});
+
+console.log(response.content);
+```
+
+## Streaming
+
+```typescript
+const client = new UnifiedLLMClient(MODELS.GPT_4O_MINI);
+
+const response = await client.chat({
+  message: "Tell me a short story.",
+  stream: true,
+  onChunk: (chunk) => {
+    process.stdout.write(chunk.content ?? "");
+  },
+  onFinalResponse: (final) => {
+    console.log("\n\nTotal tokens:", final.usage?.totalTokens);
+  },
+});
+```
+
+## Conversation History
+
+The client maintains conversation history internally. Use `getHistory()` and `setHistory()` to persist or restore it:
+
+```typescript
+const client = new UnifiedLLMClient(MODELS.CLAUDE_SONNET_4_6);
+
+await client.chat({ message: "My name is Alice." });
+await client.chat({ message: "What's my name?" });
+// → "Your name is Alice."
+
+const history = client.getHistory();
+// Save and restore later:
+const newClient = new UnifiedLLMClient(MODELS.CLAUDE_SONNET_4_6);
+newClient.setHistory(history);
+```
+
+## Multimodal Inputs
+
+Attach images or audio files directly to a message:
+
+```typescript
+const response = await client.chat({
+  message: {
+    role: "user",
+    content: "What do you see in this image?",
+    attachedFile: "./photo.jpg",
+  },
+});
+```
+
+## Audio Generation (Gemini)
+
+```typescript
+import { UnifiedLLMClient, MODELS } from "@imenam/unified-llm-client";
+
+const client = new UnifiedLLMClient(MODELS.GEMINI_FLASH_2_5);
+
+const audio = await client.generateAudio({
+  text: "Hello, this is a test of text-to-speech.",
+  outputPath: "./output.wav",
+  voiceConfig: { voiceName: "Kore" },
+});
+```
+
+## Logging
+
+Pass a `logger` callback to capture structured logs for every request:
+
+```typescript
+const client = new UnifiedLLMClient(MODELS.GPT_4O_MINI, {
+  logger: (log) => {
+    console.log(`[${log.provider}] ${log.model} — ${log.totalTokens} tokens`);
+  },
+});
+```
+
+## API Reference
+
+### `new UnifiedLLMClient(model, options?)`
+
+| Parameter | Type | Description |
+|-----------|------|-------------|
+| `model` | `Model` | Model object from `MODELS` or a custom `{ provider, modelName }` |
+| `options.logger` | `(log: UnifiedLog) => void` | Optional logging callback |
+
+### `client.chat(options)`
+
+| Option | Type | Default | Description |
+|--------|------|---------|-------------|
+| `message` | `string \| ChatMessage \| ChatMessage[]` | — | User message or history |
+| `tools` | `ToolDefinition[]` | `[]` | Tools available to the LLM |
+| `toolChoice` | `"auto" \| "none" \| string` | `"auto"` | Tool selection strategy |
+| `stream` | `boolean` | `false` | Enable streaming |
+| `onChunk` | `(chunk: StreamedChatChunk) => void` | — | Streaming chunk callback |
+| `onFinalResponse` | `(res: ChatResponse) => void` | — | Final response callback |
+
+### `ChatResponse`
+
+```typescript
+interface ChatResponse {
+  content: string | null;
+  toolCalls?: ToolCall[];
+  usage?: {
+    promptTokens: number;
+    completionTokens: number;
+    totalTokens: number;
+  };
+  raw: unknown;
+}
+```
+
+## License
+
+MIT

package/dist/adapters/anthropic/AnthropicLogAdapter.d.ts

@@ -0,0 +1,6 @@
+import { UnifiedLog } from "../../common/types";
+import { LogAdapter, LogData } from "../../core/LLMAdapter";
+export declare class AnthropicLogAdapter implements LogAdapter {
+    toUnifiedLog(data: LogData): UnifiedLog;
+}
+//# sourceMappingURL=AnthropicLogAdapter.d.ts.map

package/dist/adapters/anthropic/AnthropicLogAdapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"AnthropicLogAdapter.d.ts","sourceRoot":"","sources":["../../../src/adapters/anthropic/AnthropicLogAdapter.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAY,MAAM,oBAAoB,CAAC;AAC1D,OAAO,EAAE,UAAU,EAAE,OAAO,EAAE,MAAM,uBAAuB,CAAC;AAE5D,qBAAa,mBAAoB,YAAW,UAAU;IACpD,YAAY,CAAC,IAAI,EAAE,OAAO,GAAG,UAAU;CAoCxC"}

package/dist/adapters/anthropic/AnthropicLogAdapter.js

@@ -0,0 +1,41 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.AnthropicLogAdapter = void 0;
+class AnthropicLogAdapter {
+    toUnifiedLog(data) {
+        var _a, _b, _c, _d, _e, _f, _g, _h, _j, _k, _l;
+        const { model, provider, messages, response, latencyMs, metadata } = data;
+        const usage = response.usage;
+        const raw = response.raw;
+        const inputTokens = (_c = (_b = (_a = raw === null || raw === void 0 ? void 0 : raw.usage) === null || _a === void 0 ? void 0 : _a.input_tokens) !== null && _b !== void 0 ? _b : usage === null || usage === void 0 ? void 0 : usage.promptTokens) !== null && _c !== void 0 ? _c : 0;
+        const outputTokens = (_f = (_e = (_d = raw === null || raw === void 0 ? void 0 : raw.usage) === null || _d === void 0 ? void 0 : _d.output_tokens) !== null && _e !== void 0 ? _e : usage === null || usage === void 0 ? void 0 : usage.completionTokens) !== null && _f !== void 0 ? _f : 0;
+        const cacheCreationTokens = (_h = (_g = raw === null || raw === void 0 ? void 0 : raw.usage) === null || _g === void 0 ? void 0 : _g.cache_creation_input_tokens) !== null && _h !== void 0 ? _h : 0;
+        const cacheReadTokens = (_k = (_j = raw === null || raw === void 0 ? void 0 : raw.usage) === null || _j === void 0 ? void 0 : _j.cache_read_input_tokens) !== null && _k !== void 0 ? _k : 0;
+        return {
+            requestId: (_l = raw === null || raw === void 0 ? void 0 : raw.id) !== null && _l !== void 0 ? _l : `anthropic-${Date.now()}-${Math.random().toString(36).substr(2, 9)}`,
+            timestamp: new Date(),
+            provider: provider,
+            model,
+            prompt: messages,
+            output: {
+                content: response.content,
+                toolCalls: response.toolCalls,
+                finishReason: response.finishReason,
+            },
+            usage: {
+                promptTokens: inputTokens,
+                completionTokens: outputTokens,
+                totalTokens: inputTokens + outputTokens,
+                promptDetails: {
+                    cachedTokens: cacheCreationTokens + cacheReadTokens,
+                },
+                completionDetails: {
+                    textTokens: outputTokens,
+                },
+            },
+            latencyMs,
+            metadata,
+        };
+    }
+}
+exports.AnthropicLogAdapter = AnthropicLogAdapter;

package/dist/adapters/anthropic/AnthropicOutputAdapter.d.ts

@@ -0,0 +1,6 @@
+import { ChatResponse } from "../../common/types";
+import { OutputAdapter } from "../../core/LLMAdapter";
+export declare class AnthropicOutputAdapter implements OutputAdapter {
+    parse(providerResponse: any): ChatResponse;
+}
+//# sourceMappingURL=AnthropicOutputAdapter.d.ts.map

package/dist/adapters/anthropic/AnthropicOutputAdapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"AnthropicOutputAdapter.d.ts","sourceRoot":"","sources":["../../../src/adapters/anthropic/AnthropicOutputAdapter.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,YAAY,EAAY,MAAM,oBAAoB,CAAC;AAC5D,OAAO,EAAE,aAAa,EAAE,MAAM,uBAAuB,CAAC;AAEtD,qBAAa,sBAAuB,YAAW,aAAa;IAC1D,KAAK,CAAC,gBAAgB,EAAE,GAAG,GAAG,YAAY;CAqC3C"}

package/dist/adapters/anthropic/AnthropicOutputAdapter.js

@@ -0,0 +1,41 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.AnthropicOutputAdapter = void 0;
+class AnthropicOutputAdapter {
+    parse(providerResponse) {
+        var _a, _b, _c, _d;
+        const raw = (_a = providerResponse.raw) !== null && _a !== void 0 ? _a : providerResponse;
+        let textContent = null;
+        const toolCalls = [];
+        for (const block of (_b = raw.content) !== null && _b !== void 0 ? _b : []) {
+            if (block.type === "text") {
+                textContent = (textContent !== null && textContent !== void 0 ? textContent : "") + block.text;
+            }
+            else if (block.type === "tool_use") {
+                toolCalls.push({
+                    id: block.id,
+                    name: block.name,
+                    arguments: block.input,
+                });
+            }
+        }
+        const usage = raw.usage;
+        return {
+            content: textContent,
+            toolCalls: toolCalls.length > 0 ? toolCalls : undefined,
+            lastToolArguments: (_c = toolCalls[0]) === null || _c === void 0 ? void 0 : _c.arguments,
+            raw,
+            usage: usage
+                ? {
+                    promptTokens: usage.input_tokens,
+                    completionTokens: usage.output_tokens,
+                    totalTokens: usage.input_tokens + usage.output_tokens,
+                    promptTokensDetails: null,
+                    thoughtsToken: 0,
+                }
+                : undefined,
+            finishReason: (_d = raw.stop_reason) !== null && _d !== void 0 ? _d : undefined,
+        };
+    }
+}
+exports.AnthropicOutputAdapter = AnthropicOutputAdapter;

package/dist/adapters/anthropic/AnthropicPromptAdapter.d.ts

@@ -0,0 +1,12 @@
+import { ChatMessage, ChatResponse, ToolDefinition } from "../../common/types";
+import { ChatOptionalConfig, ChatStreamOptions, PromptAdapter, PromptAdapterOptions } from "../../core/LLMAdapter";
+export declare class AnthropicPromptAdapter implements PromptAdapter {
+    private client;
+    private opts;
+    constructor(opts: PromptAdapterOptions);
+    formatMessage(message: ChatMessage): Promise<any>;
+    private buildRequestParams;
+    sendToProvider(messages: any[], tools?: ToolDefinition[], toolChoice?: "auto" | "none" | string, _attachedFile?: string | string[], model?: string, _config?: ChatOptionalConfig): Promise<ChatResponse>;
+    streamFromProvider({ messages, tools, toolChoice, model, onChunk, onFinalResponse, }: ChatStreamOptions): Promise<ChatResponse>;
+}
+//# sourceMappingURL=AnthropicPromptAdapter.d.ts.map

package/dist/adapters/anthropic/AnthropicPromptAdapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"AnthropicPromptAdapter.d.ts","sourceRoot":"","sources":["../../../src/adapters/anthropic/AnthropicPromptAdapter.ts"],"names":[],"mappings":"AAIA,OAAO,EAAE,WAAW,EAAE,YAAY,EAAY,cAAc,EAAE,MAAM,oBAAoB,CAAC;AACzF,OAAO,EAAE,kBAAkB,EAAE,iBAAiB,EAAE,aAAa,EAAE,oBAAoB,EAAqB,MAAM,uBAAuB,CAAC;AA4BtI,qBAAa,sBAAuB,YAAW,aAAa;IAC1D,OAAO,CAAC,MAAM,CAAY;IAC1B,OAAO,CAAC,IAAI,CAAuB;gBAEvB,IAAI,EAAE,oBAAoB;IAKhC,aAAa,CAAC,OAAO,EAAE,WAAW,GAAG,OAAO,CAAC,GAAG,CAAC;IAsEvD,OAAO,CAAC,kBAAkB;IAmCpB,cAAc,CAClB,QAAQ,EAAE,GAAG,EAAE,EACf,KAAK,GAAE,cAAc,EAAO,EAC5B,UAAU,GAAE,MAAM,GAAG,MAAM,GAAG,MAAe,EAC7C,aAAa,CAAC,EAAE,MAAM,GAAG,MAAM,EAAE,EACjC,KAAK,CAAC,EAAE,MAAM,EACd,OAAO,CAAC,EAAE,kBAAkB,GAC3B,OAAO,CAAC,YAAY,CAAC;IAgDlB,kBAAkB,CAAC,EACvB,QAAQ,EACR,KAAK,EACL,UAAU,EACV,KAAK,EACL,OAAO,EACP,eAAe,GAChB,EAAE,iBAAiB,GAAG,OAAO,CAAC,YAAY,CAAC;CA8D7C"}

package/dist/adapters/anthropic/AnthropicPromptAdapter.js

@@ -0,0 +1,249 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.AnthropicPromptAdapter = void 0;
+const sdk_1 = __importDefault(require("@anthropic-ai/sdk"));
+const fs_1 = __importDefault(require("fs"));
+const mime = __importStar(require("mime-types"));
+const sharp_1 = __importDefault(require("sharp"));
+const IMAGE_MIME_TYPES = ["image/jpeg", "image/png", "image/gif", "image/webp"];
+const ANTHROPIC_MAX_IMAGE_BYTES = 5 * 1024 * 1024; // 5 MB
+function compressImageIfNeeded(filePath, mimeType) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const raw = fs_1.default.readFileSync(filePath);
+        if (raw.byteLength <= ANTHROPIC_MAX_IMAGE_BYTES) {
+            return { data: raw.toString("base64"), media_type: mimeType };
+        }
+        // Compression progressive jusqu'à passer sous la limite
+        let quality = 85;
+        let compressed;
+        do {
+            compressed = yield (0, sharp_1.default)(raw).jpeg({ quality }).toBuffer();
+            quality -= 10;
+        } while (compressed.byteLength > ANTHROPIC_MAX_IMAGE_BYTES && quality > 10);
+        return { data: compressed.toString("base64"), media_type: "image/jpeg" };
+    });
+}
+class AnthropicPromptAdapter {
+    constructor(opts) {
+        this.opts = opts;
+        this.client = new sdk_1.default({ apiKey: opts.apiKey });
+    }
+    formatMessage(message) {
+        return __awaiter(this, void 0, void 0, function* () {
+            var _a, _b;
+            if (message.role === "system") {
+                return { role: "system", content: message.content };
+            }
+            // Tool result message → user turn with tool_result content block
+            if (message.role === "tool") {
+                return {
+                    role: "user",
+                    content: [{
+                            type: "tool_result",
+                            tool_use_id: (_a = message.tool_call_id) !== null && _a !== void 0 ? _a : "",
+                            content: typeof message.content === "string" ? message.content : JSON.stringify(message.content),
+                        }],
+                };
+            }
+            // Assistant message with tool calls → assistant turn with tool_use content blocks
+            if (message.role === "assistant" && message.tool_calls && message.tool_calls.length > 0) {
+                const contentBlocks = [];
+                if (message.content) {
+                    contentBlocks.push({ type: "text", text: message.content });
+                }
+                for (const tc of message.tool_calls) {
+                    contentBlocks.push({
+                        type: "tool_use",
+                        id: (_b = tc.id) !== null && _b !== void 0 ? _b : tc.name,
+                        name: tc.name,
+                        input: tc.arguments,
+                    });
+                }
+                return { role: "assistant", content: contentBlocks };
+            }
+            if (message.attachedFile) {
+                const files = Array.isArray(message.attachedFile) ? message.attachedFile : [message.attachedFile];
+                const contentBlocks = [];
+                for (const filePath of files) {
+                    const mimeType = mime.lookup(filePath) || "application/octet-stream";
+                    if (IMAGE_MIME_TYPES.includes(mimeType)) {
+                        const { data, media_type } = yield compressImageIfNeeded(filePath, mimeType);
+                        contentBlocks.push({
+                            type: "image",
+                            source: { type: "base64", media_type, data },
+                        });
+                    }
+                    else if (mimeType === "application/pdf") {
+                        const base64Data = fs_1.default.readFileSync(filePath).toString("base64");
+                        contentBlocks.push({
+                            type: "document",
+                            source: { type: "base64", media_type: "application/pdf", data: base64Data },
+                        });
+                    }
+                }
+                contentBlocks.push({ type: "text", text: message.content });
+                return {
+                    role: message.role === "assistant" ? "assistant" : "user",
+                    content: contentBlocks,
+                };
+            }
+            return {
+                role: message.role === "assistant" ? "assistant" : "user",
+                content: message.content,
+            };
+        });
+    }
+    buildRequestParams(messages, tools = [], toolChoice = "auto", model) {
+        const systemMessages = messages.filter((m) => m.role === "system");
+        const nonSystemMessages = messages.filter((m) => m.role !== "system");
+        const systemPrompt = systemMessages.map((m) => m.content).join("\n") || undefined;
+        let sdkTools = undefined;
+        let sdkToolChoice = undefined;
+        if (toolChoice !== "none" && tools.length > 0) {
+            sdkTools = tools.map((t) => {
+                var _a, _b;
+                return ({
+                    name: t.name,
+                    description: t.description,
+                    input_schema: {
+                        type: "object",
+                        properties: ((_a = t.parameters) === null || _a === void 0 ? void 0 : _a.properties) || {},
+                        required: ((_b = t.parameters) === null || _b === void 0 ? void 0 : _b.required) || [],
+                    },
+                });
+            });
+            if (toolChoice === "auto") {
+                sdkToolChoice = { type: "auto" };
+            }
+            else {
+                sdkToolChoice = { type: "tool", name: toolChoice };
+            }
+        }
+        return { systemPrompt, nonSystemMessages, sdkTools, sdkToolChoice, model: model || this.opts.model };
+    }
+    sendToProvider(messages_1) {
+        return __awaiter(this, arguments, void 0, function* (messages, tools = [], toolChoice = "auto", _attachedFile, model, _config) {
+            var _a, _b;
+            const { systemPrompt, nonSystemMessages, sdkTools, sdkToolChoice, model: resolvedModel } = this.buildRequestParams(messages, tools, toolChoice, model);
+            const response = yield this.client.messages.create(Object.assign(Object.assign(Object.assign({ model: resolvedModel, max_tokens: 8096, messages: nonSystemMessages }, (systemPrompt ? { system: systemPrompt } : {})), (sdkTools ? { tools: sdkTools } : {})), (sdkToolChoice ? { tool_choice: sdkToolChoice } : {})));
+            let textContent = null;
+            const toolCalls = [];
+            for (const block of response.content) {
+                if (block.type === "text") {
+                    textContent = (textContent !== null && textContent !== void 0 ? textContent : "") + block.text;
+                }
+                else if (block.type === "tool_use") {
+                    toolCalls.push({
+                        id: block.id,
+                        name: block.name,
+                        arguments: block.input,
+                    });
+                }
+            }
+            return {
+                content: textContent,
+                toolCalls: toolCalls.length > 0 ? toolCalls : undefined,
+                lastToolArguments: (_a = toolCalls[0]) === null || _a === void 0 ? void 0 : _a.arguments,
+                raw: response,
+                usage: {
+                    promptTokens: response.usage.input_tokens,
+                    completionTokens: response.usage.output_tokens,
+                    totalTokens: response.usage.input_tokens + response.usage.output_tokens,
+                    promptTokensDetails: null,
+                    thoughtsToken: 0,
+                },
+                finishReason: (_b = response.stop_reason) !== null && _b !== void 0 ? _b : undefined,
+            };
+        });
+    }
+    streamFromProvider(_a) {
+        return __awaiter(this, arguments, void 0, function* ({ messages, tools, toolChoice, model, onChunk, onFinalResponse, }) {
+            var _b, _c;
+            const { systemPrompt, nonSystemMessages, sdkTools, sdkToolChoice, model: resolvedModel } = this.buildRequestParams(messages, tools !== null && tools !== void 0 ? tools : [], toolChoice !== null && toolChoice !== void 0 ? toolChoice : "auto", model);
+            const stream = this.client.messages.stream(Object.assign(Object.assign(Object.assign({ model: resolvedModel, max_tokens: 8096, messages: nonSystemMessages }, (systemPrompt ? { system: systemPrompt } : {})), (sdkTools ? { tools: sdkTools } : {})), (sdkToolChoice ? { tool_choice: sdkToolChoice } : {})));
+            const toolCalls = [];
+            stream.on("text", (text) => {
+                onChunk === null || onChunk === void 0 ? void 0 : onChunk({ contentChunk: text });
+            });
+            stream.on("contentBlock", (block) => {
+                if (block.type === "tool_use") {
+                    toolCalls.push({
+                        id: block.id,
+                        name: block.name,
+                        arguments: block.input,
+                    });
+                }
+            });
+            const finalMessage = yield stream.finalMessage();
+            let textContent = null;
+            for (const block of finalMessage.content) {
+                if (block.type === "text") {
+                    textContent = (textContent !== null && textContent !== void 0 ? textContent : "") + block.text;
+                }
+            }
+            const finalResponse = {
+                content: textContent,
+                toolCalls: toolCalls.length > 0 ? toolCalls : undefined,
+                lastToolArguments: (_b = toolCalls[0]) === null || _b === void 0 ? void 0 : _b.arguments,
+                raw: finalMessage,
+                usage: {
+                    promptTokens: finalMessage.usage.input_tokens,
+                    completionTokens: finalMessage.usage.output_tokens,
+                    totalTokens: finalMessage.usage.input_tokens + finalMessage.usage.output_tokens,
+                    promptTokensDetails: null,
+                    thoughtsToken: 0,
+                },
+                finishReason: (_c = finalMessage.stop_reason) !== null && _c !== void 0 ? _c : undefined,
+            };
+            onChunk === null || onChunk === void 0 ? void 0 : onChunk({ isFinal: true, contentChunk: "" });
+            onFinalResponse === null || onFinalResponse === void 0 ? void 0 : onFinalResponse(finalResponse);
+            return finalResponse;
+        });
+    }
+}
+exports.AnthropicPromptAdapter = AnthropicPromptAdapter;

package/dist/adapters/anthropic/AnthropicToolAdapter.d.ts

@@ -0,0 +1,6 @@
+import { ToolCall } from "../../common/types";
+import { ToolAdapter } from "../../core/LLMAdapter";
+export declare class AnthropicToolAdapter implements ToolAdapter {
+    formatToolCallResult(toolCall: ToolCall): any;
+}
+//# sourceMappingURL=AnthropicToolAdapter.d.ts.map

package/dist/adapters/anthropic/AnthropicToolAdapter.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"AnthropicToolAdapter.d.ts","sourceRoot":"","sources":["../../../src/adapters/anthropic/AnthropicToolAdapter.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,WAAW,EAAE,MAAM,uBAAuB,CAAC;AAEpD,qBAAa,oBAAqB,YAAW,WAAW;IACtD,oBAAoB,CAAC,QAAQ,EAAE,QAAQ,GAAG,GAAG;CAe9C"}

package/dist/adapters/anthropic/AnthropicToolAdapter.js

@@ -0,0 +1,20 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.AnthropicToolAdapter = void 0;
+class AnthropicToolAdapter {
+    formatToolCallResult(toolCall) {
+        return {
+            role: "user",
+            content: [
+                {
+                    type: "tool_result",
+                    tool_use_id: toolCall.id,
+                    content: typeof toolCall.result === "string"
+                        ? toolCall.result
+                        : JSON.stringify(toolCall.result),
+                },
+            ],
+        };
+    }
+}
+exports.AnthropicToolAdapter = AnthropicToolAdapter;