@semiont/inference 0.5.2 → 0.5.4

package/dist/factory.d.ts

@@ -0,0 +1,12 @@
+import type { Logger } from '@semiont/core';
+import { InferenceClient } from './interface.js';
+export type InferenceClientType = 'anthropic' | 'ollama';
+export interface InferenceClientConfig {
+    type: InferenceClientType;
+    apiKey?: string;
+    model: string;
+    endpoint?: string;
+    baseURL?: string;
+}
+export declare function createInferenceClient(config: InferenceClientConfig, logger?: Logger): InferenceClient;
+//# sourceMappingURL=factory.d.ts.map

package/dist/factory.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"factory.d.ts","sourceRoot":"","sources":["../src/factory.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,eAAe,CAAC;AAC5C,OAAO,EAAE,eAAe,EAAE,MAAM,gBAAgB,CAAC;AAIjD,MAAM,MAAM,mBAAmB,GAAG,WAAW,GAAG,QAAQ,CAAC;AAEzD,MAAM,WAAW,qBAAqB;IACpC,IAAI,EAAE,mBAAmB,CAAC;IAC1B,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,wBAAgB,qBAAqB,CAAC,MAAM,EAAE,qBAAqB,EAAE,MAAM,CAAC,EAAE,MAAM,GAAG,eAAe,CAyBrG"}

package/dist/implementations/anthropic.d.ts

@@ -0,0 +1,12 @@
+import type { Logger } from '@semiont/core';
+import { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';
+export declare class AnthropicInferenceClient implements InferenceClient {
+    readonly type: "anthropic";
+    readonly modelId: string;
+    private client;
+    private logger?;
+    constructor(apiKey: string, model: string, baseURL?: string, logger?: Logger);
+    generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string>;
+    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse>;
+}
+//# sourceMappingURL=anthropic.d.ts.map

package/dist/implementations/anthropic.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"anthropic.d.ts","sourceRoot":"","sources":["../../src/implementations/anthropic.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,eAAe,CAAC;AAE5C,OAAO,EAAE,eAAe,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,iBAAiB,CAAC;AAEvF,qBAAa,wBAAyB,YAAW,eAAe;IAC9D,QAAQ,CAAC,IAAI,EAAG,WAAW,CAAU;IACrC,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,OAAO,CAAC,MAAM,CAAY;IAC1B,OAAO,CAAC,MAAM,CAAC,CAAS;gBAEZ,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,MAAM;IAStE,YAAY,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,MAAM,CAAC;IAKjH,wBAAwB,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,iBAAiB,CAAC;CAqG/I"}

package/dist/implementations/mock.d.ts

@@ -0,0 +1,20 @@
+import { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';
+export declare class MockInferenceClient implements InferenceClient {
+    readonly type: "mock";
+    readonly modelId: "mock-model";
+    private responses;
+    private responseIndex;
+    private stopReasons;
+    calls: Array<{
+        prompt: string;
+        maxTokens: number;
+        temperature: number;
+        options?: InferenceOptions;
+    }>;
+    constructor(responses?: string[], stopReasons?: string[]);
+    generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string>;
+    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse>;
+    reset(): void;
+    setResponses(responses: string[], stopReasons?: string[]): void;
+}
+//# sourceMappingURL=mock.d.ts.map

package/dist/implementations/mock.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"mock.d.ts","sourceRoot":"","sources":["../../src/implementations/mock.ts"],"names":[],"mappings":"AAEA,OAAO,EAAE,eAAe,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,iBAAiB,CAAC;AAEvF,qBAAa,mBAAoB,YAAW,eAAe;IACzD,QAAQ,CAAC,IAAI,EAAG,MAAM,CAAU;IAChC,QAAQ,CAAC,OAAO,EAAG,YAAY,CAAU;IACzC,OAAO,CAAC,SAAS,CAAgB;IACjC,OAAO,CAAC,aAAa,CAAa;IAClC,OAAO,CAAC,WAAW,CAAgB;IAC5B,KAAK,EAAE,KAAK,CAAC;QAAE,MAAM,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAC;QAAC,WAAW,EAAE,MAAM,CAAC;QAAC,OAAO,CAAC,EAAE,gBAAgB,CAAA;KAAE,CAAC,CAAM;gBAErG,SAAS,GAAE,MAAM,EAAsB,EAAE,WAAW,CAAC,EAAE,MAAM,EAAE;IAKrE,YAAY,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,MAAM,CAAC;IAKjH,wBAAwB,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,iBAAiB,CAAC;IAc9I,KAAK,IAAI,IAAI;IAKb,YAAY,CAAC,SAAS,EAAE,MAAM,EAAE,EAAE,WAAW,CAAC,EAAE,MAAM,EAAE,GAAG,IAAI;CAKhE"}

package/dist/implementations/ollama.d.ts

@@ -0,0 +1,12 @@
+import type { Logger } from '@semiont/core';
+import { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';
+export declare class OllamaInferenceClient implements InferenceClient {
+    readonly type: "ollama";
+    readonly modelId: string;
+    private baseURL;
+    private logger?;
+    constructor(model: string, baseURL?: string, logger?: Logger);
+    generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string>;
+    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse>;
+}
+//# sourceMappingURL=ollama.d.ts.map

package/dist/implementations/ollama.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"ollama.d.ts","sourceRoot":"","sources":["../../src/implementations/ollama.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,eAAe,CAAC;AAE5C,OAAO,EAAE,eAAe,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,iBAAiB,CAAC;AAYvF,qBAAa,qBAAsB,YAAW,eAAe;IAC3D,QAAQ,CAAC,IAAI,EAAG,QAAQ,CAAU;IAClC,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,OAAO,CAAC,OAAO,CAAS;IACxB,OAAO,CAAC,MAAM,CAAC,CAAS;gBAEZ,KAAK,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,EAAE,MAAM;IAMtD,YAAY,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,MAAM,CAAC;IAKjH,wBAAwB,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,iBAAiB,CAAC;CA0G/I"}

package/dist/index.d.ts

@@ -1,78 +1,6 @@
-import { Logger } from '@semiont/core';
-
-interface InferenceResponse {
-    text: string;
-    stopReason: 'end_turn' | 'max_tokens' | 'stop_sequence' | string;
-}
-interface InferenceClient {
-    /** Provider type identifier (e.g. 'anthropic', 'ollama') */
-    readonly type: string;
-    /** Model identifier used for generation (e.g. 'claude-opus-4-6', 'llama3') */
-    readonly modelId: string;
-    /**
-     * Generate text from a prompt (simple interface)
-     * @param prompt - The input prompt
-     * @param maxTokens - Maximum tokens to generate
-     * @param temperature - Sampling temperature (0-1)
-     * @returns Generated text
-     */
-    generateText(prompt: string, maxTokens: number, temperature: number): Promise<string>;
-    /**
-     * Generate text with detailed response information
-     * @param prompt - The input prompt
-     * @param maxTokens - Maximum tokens to generate
-     * @param temperature - Sampling temperature (0-1)
-     * @returns Response with text and metadata
-     */
-    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse>;
-}
-
-type InferenceClientType = 'anthropic' | 'ollama';
-interface InferenceClientConfig {
-    type: InferenceClientType;
-    apiKey?: string;
-    model: string;
-    endpoint?: string;
-    baseURL?: string;
-}
-declare function createInferenceClient(config: InferenceClientConfig, logger?: Logger): InferenceClient;
-
-declare class AnthropicInferenceClient implements InferenceClient {
-    readonly type: "anthropic";
-    readonly modelId: string;
-    private client;
-    private logger?;
-    constructor(apiKey: string, model: string, baseURL?: string, logger?: Logger);
-    generateText(prompt: string, maxTokens: number, temperature: number): Promise<string>;
-    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse>;
-}
-
-declare class OllamaInferenceClient implements InferenceClient {
-    readonly type: "ollama";
-    readonly modelId: string;
-    private baseURL;
-    private logger?;
-    constructor(model: string, baseURL?: string, logger?: Logger);
-    generateText(prompt: string, maxTokens: number, temperature: number): Promise<string>;
-    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse>;
-}
-
-declare class MockInferenceClient implements InferenceClient {
-    readonly type: "mock";
-    readonly modelId: "mock-model";
-    private responses;
-    private responseIndex;
-    private stopReasons;
-    calls: Array<{
-        prompt: string;
-        maxTokens: number;
-        temperature: number;
-    }>;
-    constructor(responses?: string[], stopReasons?: string[]);
-    generateText(prompt: string, maxTokens: number, temperature: number): Promise<string>;
-    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse>;
-    reset(): void;
-    setResponses(responses: string[], stopReasons?: string[]): void;
-}
-
-export { AnthropicInferenceClient, type InferenceClient, type InferenceClientConfig, type InferenceClientType, type InferenceResponse, MockInferenceClient, OllamaInferenceClient, createInferenceClient };
+export { createInferenceClient, type InferenceClientConfig, type InferenceClientType, } from './factory';
+export { type InferenceClient, type InferenceOptions, type InferenceResponse } from './interface';
+export { AnthropicInferenceClient } from './implementations/anthropic';
+export { OllamaInferenceClient } from './implementations/ollama';
+export { MockInferenceClient } from './implementations/mock';
+//# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AACA,OAAO,EACL,qBAAqB,EACrB,KAAK,qBAAqB,EAC1B,KAAK,mBAAmB,GACzB,MAAM,WAAW,CAAC;AAEnB,OAAO,EAAE,KAAK,eAAe,EAAE,KAAK,gBAAgB,EAAE,KAAK,iBAAiB,EAAE,MAAM,aAAa,CAAC;AAClG,OAAO,EAAE,wBAAwB,EAAE,MAAM,6BAA6B,CAAC;AACvE,OAAO,EAAE,qBAAqB,EAAE,MAAM,0BAA0B,CAAC;AACjE,OAAO,EAAE,mBAAmB,EAAE,MAAM,wBAAwB,CAAC"}

package/dist/index.js

@@ -14,17 +14,26 @@ var AnthropicInferenceClient = class {
     this.modelId = model;
     this.logger = logger;
   }
-  async generateText(prompt, maxTokens, temperature) {
-    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);
+  async generateText(prompt, maxTokens, temperature, options) {
+    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);
     return response.text;
   }
-  async generateTextWithMetadata(prompt, maxTokens, temperature) {
+  async generateTextWithMetadata(prompt, maxTokens, temperature, options) {
+    const jsonMode = options?.format === "json";
+    const prefill = jsonMode ? "[" : void 0;
     this.logger?.debug("Generating text with inference client", {
       model: this.modelId,
       promptLength: prompt.length,
       maxTokens,
-      temperature
+      temperature,
+      format: options?.format
     });
+    const messages = [
+      { role: "user", content: prompt }
+    ];
+    if (prefill) {
+      messages.push({ role: "assistant", content: prefill });
+    }
     const start = performance.now();
     let response;
     try {
@@ -32,12 +41,7 @@ var AnthropicInferenceClient = class {
         model: this.modelId,
         max_tokens: maxTokens,
         temperature,
-        messages: [
-          {
-            role: "user",
-            content: prompt
-          }
-        ]
+        messages
       });
     } catch (err) {
       recordInferenceUsage({
@@ -77,13 +81,14 @@ var AnthropicInferenceClient = class {
       inputTokens: response.usage?.input_tokens,
       outputTokens: response.usage?.output_tokens
     });
+    const text = prefill ? prefill + textContent.text : textContent.text;
     this.logger?.info("Text generation completed", {
       model: this.modelId,
-      textLength: textContent.text.length,
+      textLength: text.length,
       stopReason: response.stop_reason
     });
     return {
-      text: textContent.text,
+      text,
       stopReason: response.stop_reason || "unknown"
     };
   }
@@ -101,34 +106,39 @@ var OllamaInferenceClient = class {
     this.modelId = model;
     this.logger = logger;
   }
-  async generateText(prompt, maxTokens, temperature) {
-    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);
+  async generateText(prompt, maxTokens, temperature, options) {
+    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);
     return response.text;
   }
-  async generateTextWithMetadata(prompt, maxTokens, temperature) {
+  async generateTextWithMetadata(prompt, maxTokens, temperature, options) {
     this.logger?.debug("Generating text with Ollama", {
       model: this.modelId,
       promptLength: prompt.length,
       maxTokens,
-      temperature
+      temperature,
+      format: options?.format
     });
     const url = `${this.baseURL}/api/generate`;
     const start = performance.now();
+    const body = {
+      model: this.modelId,
+      prompt,
+      stream: false,
+      think: false,
+      options: {
+        num_predict: maxTokens,
+        temperature
+      }
+    };
+    if (options?.format === "json") {
+      body["format"] = { type: "array", items: {} };
+    }
     let res;
     try {
       res = await fetch(url, {
         method: "POST",
         headers: { "Content-Type": "application/json" },
-        body: JSON.stringify({
-          model: this.modelId,
-          prompt,
-          stream: false,
-          think: false,
-          options: {
-            num_predict: maxTokens,
-            temperature
-          }
-        })
+        body: JSON.stringify(body)
       });
     } catch (err) {
       recordInferenceUsage2({
@@ -146,13 +156,13 @@ var OllamaInferenceClient = class {
         durationMs: performance.now() - start,
         outcome: "error"
       });
-      const body = await res.text();
+      const body2 = await res.text();
       this.logger?.error("Ollama API error", {
         model: this.modelId,
         status: res.status,
-        body
+        body: body2
       });
-      throw new Error(`Ollama API error (${res.status}): ${body}`);
+      throw new Error(`Ollama API error (${res.status}): ${body2}`);
     }
     const data = await res.json();
     if (!data.response) {
@@ -236,12 +246,12 @@ var MockInferenceClient = class {
     this.responses = responses;
     this.stopReasons = stopReasons || responses.map(() => "end_turn");
   }
-  async generateText(prompt, maxTokens, temperature) {
-    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);
+  async generateText(prompt, maxTokens, temperature, options) {
+    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);
     return response.text;
   }
-  async generateTextWithMetadata(prompt, maxTokens, temperature) {
-    this.calls.push({ prompt, maxTokens, temperature });
+  async generateTextWithMetadata(prompt, maxTokens, temperature, options) {
+    this.calls.push({ prompt, maxTokens, temperature, ...options ? { options } : {} });
     const text = this.responses[this.responseIndex];
     const stopReason = this.stopReasons[this.responseIndex] || "end_turn";
     if (this.responseIndex < this.responses.length - 1) {

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["../src/implementations/anthropic.ts","../src/implementations/ollama.ts","../src/factory.ts","../src/implementations/mock.ts"],"sourcesContent":["// Anthropic Claude implementation of InferenceClient interface\n\nimport Anthropic from '@anthropic-ai/sdk';\nimport type { Logger } from '@semiont/core';\nimport { recordInferenceUsage } from '@semiont/observability';\nimport { InferenceClient, InferenceResponse } from '../interface.js';\n\nexport class AnthropicInferenceClient implements InferenceClient {\n  readonly type = 'anthropic' as const;\n  readonly modelId: string;\n  private client: Anthropic;\n  private logger?: Logger;\n\n  constructor(apiKey: string, model: string, baseURL?: string, logger?: Logger) {\n    this.client = new Anthropic({\n      apiKey,\n      baseURL: baseURL || 'https://api.anthropic.com',\n    });\n    this.modelId = model;\n    this.logger = logger;\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse> {\n    this.logger?.debug('Generating text with inference client', {\n      model: this.modelId,\n      promptLength: prompt.length,\n      maxTokens,\n      temperature\n    });\n\n    const start = performance.now();\n    let response: Awaited<ReturnType<typeof this.client.messages.create>>;\n    try {\n      response = await this.client.messages.create({\n        model: this.modelId,\n        max_tokens: maxTokens,\n        temperature,\n        messages: [\n          {\n            role: 'user',\n            content: prompt\n          }\n        ]\n      });\n    } catch (err) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      throw err;\n    }\n\n    this.logger?.debug('Inference response received', {\n      model: this.modelId,\n      contentBlocks: response.content.length,\n      stopReason: response.stop_reason\n    });\n\n    const textContent = response.content.find(c => c.type === 'text');\n\n    if (!textContent || textContent.type !== 'text') {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n        inputTokens: response.usage?.input_tokens,\n        outputTokens: response.usage?.output_tokens,\n      });\n      this.logger?.error('No text content in inference response', {\n        model: this.modelId,\n        contentTypes: response.content.map(c => c.type)\n      });\n      throw new Error('No text content in inference response');\n    }\n\n    recordInferenceUsage({\n      provider: this.type,\n      model: this.modelId,\n      durationMs: performance.now() - start,\n      outcome: 'success',\n      inputTokens: response.usage?.input_tokens,\n      outputTokens: response.usage?.output_tokens,\n    });\n\n    this.logger?.info('Text generation completed', {\n      model: this.modelId,\n      textLength: textContent.text.length,\n      stopReason: response.stop_reason\n    });\n\n    return {\n      text: textContent.text,\n      stopReason: response.stop_reason || 'unknown'\n    };\n  }\n}\n","// Ollama implementation of InferenceClient interface\n// Uses native Ollama HTTP API (no SDK dependency)\n\nimport type { Logger } from '@semiont/core';\nimport { recordInferenceUsage } from '@semiont/observability';\nimport { InferenceClient, InferenceResponse } from '../interface.js';\n\ninterface OllamaGenerateResponse {\n  response: string;\n  done: boolean;\n  done_reason?: string;\n  /** Number of prompt tokens evaluated. Available on most Ollama versions. */\n  prompt_eval_count?: number;\n  /** Number of tokens generated. */\n  eval_count?: number;\n}\n\nexport class OllamaInferenceClient implements InferenceClient {\n  readonly type = 'ollama' as const;\n  readonly modelId: string;\n  private baseURL: string;\n  private logger?: Logger;\n\n  constructor(model: string, baseURL?: string, logger?: Logger) {\n    this.baseURL = (baseURL || 'http://localhost:11434').replace(/\\/+$/, '');\n    this.modelId = model;\n    this.logger = logger;\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse> {\n    this.logger?.debug('Generating text with Ollama', {\n      model: this.modelId,\n      promptLength: prompt.length,\n      maxTokens,\n      temperature\n    });\n\n    const url = `${this.baseURL}/api/generate`;\n    const start = performance.now();\n\n    let res: Response;\n    try {\n      res = await fetch(url, {\n        method: 'POST',\n        headers: { 'Content-Type': 'application/json' },\n        body: JSON.stringify({\n          model: this.modelId,\n          prompt,\n          stream: false,\n          think: false,\n          options: {\n            num_predict: maxTokens,\n            temperature,\n          },\n        }),\n      });\n    } catch (err) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      throw err;\n    }\n\n    if (!res.ok) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      const body = await res.text();\n      this.logger?.error('Ollama API error', {\n        model: this.modelId,\n        status: res.status,\n        body,\n      });\n      throw new Error(`Ollama API error (${res.status}): ${body}`);\n    }\n\n    const data = await res.json() as OllamaGenerateResponse;\n\n    if (!data.response) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n        inputTokens: data.prompt_eval_count,\n        outputTokens: data.eval_count,\n      });\n      this.logger?.error('Empty response from Ollama', { model: this.modelId });\n      throw new Error('Empty response from Ollama');\n    }\n\n    recordInferenceUsage({\n      provider: this.type,\n      model: this.modelId,\n      durationMs: performance.now() - start,\n      outcome: 'success',\n      inputTokens: data.prompt_eval_count,\n      outputTokens: data.eval_count,\n    });\n\n    const stopReason = mapStopReason(data.done_reason);\n\n    this.logger?.info('Text generation completed', {\n      model: this.modelId,\n      textLength: data.response.length,\n      stopReason,\n    });\n\n    return {\n      text: data.response,\n      stopReason,\n    };\n  }\n}\n\nfunction mapStopReason(doneReason: string | undefined): string {\n  switch (doneReason) {\n    case 'stop': return 'end_turn';\n    case 'length': return 'max_tokens';\n    default: return doneReason || 'unknown';\n  }\n}\n","// Factory for creating inference client instances based on configuration\n\nimport type { Logger } from '@semiont/core';\nimport { InferenceClient } from './interface.js';\nimport { AnthropicInferenceClient } from './implementations/anthropic.js';\nimport { OllamaInferenceClient } from './implementations/ollama.js';\n\nexport type InferenceClientType = 'anthropic' | 'ollama';\n\nexport interface InferenceClientConfig {\n  type: InferenceClientType;\n  apiKey?: string;\n  model: string;\n  endpoint?: string;\n  baseURL?: string;\n}\n\nexport function createInferenceClient(config: InferenceClientConfig, logger?: Logger): InferenceClient {\n  switch (config.type) {\n    case 'anthropic': {\n      if (!config.apiKey || config.apiKey.trim() === '') {\n        throw new Error('apiKey is required for Anthropic inference client');\n      }\n      return new AnthropicInferenceClient(\n        config.apiKey,\n        config.model,\n        config.endpoint || config.baseURL,\n        logger\n      );\n    }\n\n    case 'ollama': {\n      return new OllamaInferenceClient(\n        config.model,\n        config.endpoint || config.baseURL,\n        logger\n      );\n    }\n\n    default:\n      throw new Error(`Unsupported inference client type: ${config.type}`);\n  }\n}\n","// Mock implementation of InferenceClient for testing\n\nimport { InferenceClient, InferenceResponse } from '../interface.js';\n\nexport class MockInferenceClient implements InferenceClient {\n  readonly type = 'mock' as const;\n  readonly modelId = 'mock-model' as const;\n  private responses: string[] = [];\n  private responseIndex: number = 0;\n  private stopReasons: string[] = [];\n  public calls: Array<{ prompt: string; maxTokens: number; temperature: number }> = [];\n\n  constructor(responses: string[] = ['Mock response'], stopReasons?: string[]) {\n    this.responses = responses;\n    this.stopReasons = stopReasons || responses.map(() => 'end_turn');\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number): Promise<InferenceResponse> {\n    this.calls.push({ prompt, maxTokens, temperature });\n\n    const text = this.responses[this.responseIndex];\n    const stopReason = this.stopReasons[this.responseIndex] || 'end_turn';\n\n    if (this.responseIndex < this.responses.length - 1) {\n      this.responseIndex++;\n    }\n\n    return { text, stopReason };\n  }\n\n  // Test helper methods\n  reset(): void {\n    this.calls = [];\n    this.responseIndex = 0;\n  }\n\n  setResponses(responses: string[], stopReasons?: string[]): void {\n    this.responses = responses;\n    this.stopReasons = stopReasons || responses.map(() => 'end_turn');\n    this.responseIndex = 0;\n  }\n}\n"],"mappings":";AAEA,OAAO,eAAe;AAEtB,SAAS,4BAA4B;AAG9B,IAAM,2BAAN,MAA0D;AAAA,EACtD,OAAO;AAAA,EACP;AAAA,EACD;AAAA,EACA;AAAA,EAER,YAAY,QAAgB,OAAe,SAAkB,QAAiB;AAC5E,SAAK,SAAS,IAAI,UAAU;AAAA,MAC1B;AAAA,MACA,SAAS,WAAW;AAAA,IACtB,CAAC;AACD,SAAK,UAAU;AACf,SAAK,SAAS;AAAA,EAChB;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAsC;AAC1F,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,WAAW;AACnF,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAiD;AACjH,SAAK,QAAQ,MAAM,yCAAyC;AAAA,MAC1D,OAAO,KAAK;AAAA,MACZ,cAAc,OAAO;AAAA,MACrB;AAAA,MACA;AAAA,IACF,CAAC;AAED,UAAM,QAAQ,YAAY,IAAI;AAC9B,QAAI;AACJ,QAAI;AACF,iBAAW,MAAM,KAAK,OAAO,SAAS,OAAO;AAAA,QAC3C,OAAO,KAAK;AAAA,QACZ,YAAY;AAAA,QACZ;AAAA,QACA,UAAU;AAAA,UACR;AAAA,YACE,MAAM;AAAA,YACN,SAAS;AAAA,UACX;AAAA,QACF;AAAA,MACF,CAAC;AAAA,IACH,SAAS,KAAK;AACZ,2BAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAM;AAAA,IACR;AAEA,SAAK,QAAQ,MAAM,+BAA+B;AAAA,MAChD,OAAO,KAAK;AAAA,MACZ,eAAe,SAAS,QAAQ;AAAA,MAChC,YAAY,SAAS;AAAA,IACvB,CAAC;AAED,UAAM,cAAc,SAAS,QAAQ,KAAK,OAAK,EAAE,SAAS,MAAM;AAEhE,QAAI,CAAC,eAAe,YAAY,SAAS,QAAQ;AAC/C,2BAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,QACT,aAAa,SAAS,OAAO;AAAA,QAC7B,cAAc,SAAS,OAAO;AAAA,MAChC,CAAC;AACD,WAAK,QAAQ,MAAM,yCAAyC;AAAA,QAC1D,OAAO,KAAK;AAAA,QACZ,cAAc,SAAS,QAAQ,IAAI,OAAK,EAAE,IAAI;AAAA,MAChD,CAAC;AACD,YAAM,IAAI,MAAM,uCAAuC;AAAA,IACzD;AAEA,yBAAqB;AAAA,MACnB,UAAU,KAAK;AAAA,MACf,OAAO,KAAK;AAAA,MACZ,YAAY,YAAY,IAAI,IAAI;AAAA,MAChC,SAAS;AAAA,MACT,aAAa,SAAS,OAAO;AAAA,MAC7B,cAAc,SAAS,OAAO;AAAA,IAChC,CAAC;AAED,SAAK,QAAQ,KAAK,6BAA6B;AAAA,MAC7C,OAAO,KAAK;AAAA,MACZ,YAAY,YAAY,KAAK;AAAA,MAC7B,YAAY,SAAS;AAAA,IACvB,CAAC;AAED,WAAO;AAAA,MACL,MAAM,YAAY;AAAA,MAClB,YAAY,SAAS,eAAe;AAAA,IACtC;AAAA,EACF;AACF;;;ACnGA,SAAS,wBAAAA,6BAA4B;AAa9B,IAAM,wBAAN,MAAuD;AAAA,EACnD,OAAO;AAAA,EACP;AAAA,EACD;AAAA,EACA;AAAA,EAER,YAAY,OAAe,SAAkB,QAAiB;AAC5D,SAAK,WAAW,WAAW,0BAA0B,QAAQ,QAAQ,EAAE;AACvE,SAAK,UAAU;AACf,SAAK,SAAS;AAAA,EAChB;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAsC;AAC1F,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,WAAW;AACnF,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAiD;AACjH,SAAK,QAAQ,MAAM,+BAA+B;AAAA,MAChD,OAAO,KAAK;AAAA,MACZ,cAAc,OAAO;AAAA,MACrB;AAAA,MACA;AAAA,IACF,CAAC;AAED,UAAM,MAAM,GAAG,KAAK,OAAO;AAC3B,UAAM,QAAQ,YAAY,IAAI;AAE9B,QAAI;AACJ,QAAI;AACF,YAAM,MAAM,MAAM,KAAK;AAAA,QACrB,QAAQ;AAAA,QACR,SAAS,EAAE,gBAAgB,mBAAmB;AAAA,QAC9C,MAAM,KAAK,UAAU;AAAA,UACnB,OAAO,KAAK;AAAA,UACZ;AAAA,UACA,QAAQ;AAAA,UACR,OAAO;AAAA,UACP,SAAS;AAAA,YACP,aAAa;AAAA,YACb;AAAA,UACF;AAAA,QACF,CAAC;AAAA,MACH,CAAC;AAAA,IACH,SAAS,KAAK;AACZ,MAAAA,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAM;AAAA,IACR;AAEA,QAAI,CAAC,IAAI,IAAI;AACX,MAAAA,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAM,OAAO,MAAM,IAAI,KAAK;AAC5B,WAAK,QAAQ,MAAM,oBAAoB;AAAA,QACrC,OAAO,KAAK;AAAA,QACZ,QAAQ,IAAI;AAAA,QACZ;AAAA,MACF,CAAC;AACD,YAAM,IAAI,MAAM,qBAAqB,IAAI,MAAM,MAAM,IAAI,EAAE;AAAA,IAC7D;AAEA,UAAM,OAAO,MAAM,IAAI,KAAK;AAE5B,QAAI,CAAC,KAAK,UAAU;AAClB,MAAAA,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,QACT,aAAa,KAAK;AAAA,QAClB,cAAc,KAAK;AAAA,MACrB,CAAC;AACD,WAAK,QAAQ,MAAM,8BAA8B,EAAE,OAAO,KAAK,QAAQ,CAAC;AACxE,YAAM,IAAI,MAAM,4BAA4B;AAAA,IAC9C;AAEA,IAAAA,sBAAqB;AAAA,MACnB,UAAU,KAAK;AAAA,MACf,OAAO,KAAK;AAAA,MACZ,YAAY,YAAY,IAAI,IAAI;AAAA,MAChC,SAAS;AAAA,MACT,aAAa,KAAK;AAAA,MAClB,cAAc,KAAK;AAAA,IACrB,CAAC;AAED,UAAM,aAAa,cAAc,KAAK,WAAW;AAEjD,SAAK,QAAQ,KAAK,6BAA6B;AAAA,MAC7C,OAAO,KAAK;AAAA,MACZ,YAAY,KAAK,SAAS;AAAA,MAC1B;AAAA,IACF,CAAC;AAED,WAAO;AAAA,MACL,MAAM,KAAK;AAAA,MACX;AAAA,IACF;AAAA,EACF;AACF;AAEA,SAAS,cAAc,YAAwC;AAC7D,UAAQ,YAAY;AAAA,IAClB,KAAK;AAAQ,aAAO;AAAA,IACpB,KAAK;AAAU,aAAO;AAAA,IACtB;AAAS,aAAO,cAAc;AAAA,EAChC;AACF;;;ACnHO,SAAS,sBAAsB,QAA+B,QAAkC;AACrG,UAAQ,OAAO,MAAM;AAAA,IACnB,KAAK,aAAa;AAChB,UAAI,CAAC,OAAO,UAAU,OAAO,OAAO,KAAK,MAAM,IAAI;AACjD,cAAM,IAAI,MAAM,mDAAmD;AAAA,MACrE;AACA,aAAO,IAAI;AAAA,QACT,OAAO;AAAA,QACP,OAAO;AAAA,QACP,OAAO,YAAY,OAAO;AAAA,QAC1B;AAAA,MACF;AAAA,IACF;AAAA,IAEA,KAAK,UAAU;AACb,aAAO,IAAI;AAAA,QACT,OAAO;AAAA,QACP,OAAO,YAAY,OAAO;AAAA,QAC1B;AAAA,MACF;AAAA,IACF;AAAA,IAEA;AACE,YAAM,IAAI,MAAM,sCAAsC,OAAO,IAAI,EAAE;AAAA,EACvE;AACF;;;ACtCO,IAAM,sBAAN,MAAqD;AAAA,EACjD,OAAO;AAAA,EACP,UAAU;AAAA,EACX,YAAsB,CAAC;AAAA,EACvB,gBAAwB;AAAA,EACxB,cAAwB,CAAC;AAAA,EAC1B,QAA2E,CAAC;AAAA,EAEnF,YAAY,YAAsB,CAAC,eAAe,GAAG,aAAwB;AAC3E,SAAK,YAAY;AACjB,SAAK,cAAc,eAAe,UAAU,IAAI,MAAM,UAAU;AAAA,EAClE;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAsC;AAC1F,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,WAAW;AACnF,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAiD;AACjH,SAAK,MAAM,KAAK,EAAE,QAAQ,WAAW,YAAY,CAAC;AAElD,UAAM,OAAO,KAAK,UAAU,KAAK,aAAa;AAC9C,UAAM,aAAa,KAAK,YAAY,KAAK,aAAa,KAAK;AAE3D,QAAI,KAAK,gBAAgB,KAAK,UAAU,SAAS,GAAG;AAClD,WAAK;AAAA,IACP;AAEA,WAAO,EAAE,MAAM,WAAW;AAAA,EAC5B;AAAA;AAAA,EAGA,QAAc;AACZ,SAAK,QAAQ,CAAC;AACd,SAAK,gBAAgB;AAAA,EACvB;AAAA,EAEA,aAAa,WAAqB,aAA8B;AAC9D,SAAK,YAAY;AACjB,SAAK,cAAc,eAAe,UAAU,IAAI,MAAM,UAAU;AAChE,SAAK,gBAAgB;AAAA,EACvB;AACF;","names":["recordInferenceUsage"]}
+{"version":3,"sources":["../src/implementations/anthropic.ts","../src/implementations/ollama.ts","../src/factory.ts","../src/implementations/mock.ts"],"sourcesContent":["// Anthropic Claude implementation of InferenceClient interface\n\nimport Anthropic from '@anthropic-ai/sdk';\nimport type { Logger } from '@semiont/core';\nimport { recordInferenceUsage } from '@semiont/observability';\nimport { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';\n\nexport class AnthropicInferenceClient implements InferenceClient {\n  readonly type = 'anthropic' as const;\n  readonly modelId: string;\n  private client: Anthropic;\n  private logger?: Logger;\n\n  constructor(apiKey: string, model: string, baseURL?: string, logger?: Logger) {\n    this.client = new Anthropic({\n      apiKey,\n      baseURL: baseURL || 'https://api.anthropic.com',\n    });\n    this.modelId = model;\n    this.logger = logger;\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse> {\n    // Anthropic has no grammar-constrained sampling layer like Ollama's\n    // `format: \"json\"`. The closest equivalent is *prefill*: add an\n    // assistant turn whose content is the opening bracket of the\n    // expected structure. Claude continues from there, syntactically\n    // committed to producing valid JSON. The prefill characters don't\n    // appear in the response, so we prepend them ourselves on return.\n    //\n    // We assume `[` (array) — every current caller of the inference\n    // layer that asks for JSON expects an array (entity extraction,\n    // motivation detection). If an object-emitting caller appears, the\n    // option needs to grow a `root: 'array' | 'object'` field; for now\n    // a single shape keeps the contract small.\n    const jsonMode = options?.format === 'json';\n    const prefill = jsonMode ? '[' : undefined;\n\n    this.logger?.debug('Generating text with inference client', {\n      model: this.modelId,\n      promptLength: prompt.length,\n      maxTokens,\n      temperature,\n      format: options?.format,\n    });\n\n    const messages: Array<{ role: 'user' | 'assistant'; content: string }> = [\n      { role: 'user', content: prompt },\n    ];\n    if (prefill) {\n      messages.push({ role: 'assistant', content: prefill });\n    }\n\n    const start = performance.now();\n    let response: Awaited<ReturnType<typeof this.client.messages.create>>;\n    try {\n      response = await this.client.messages.create({\n        model: this.modelId,\n        max_tokens: maxTokens,\n        temperature,\n        messages,\n      });\n    } catch (err) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      throw err;\n    }\n\n    this.logger?.debug('Inference response received', {\n      model: this.modelId,\n      contentBlocks: response.content.length,\n      stopReason: response.stop_reason\n    });\n\n    const textContent = response.content.find(c => c.type === 'text');\n\n    if (!textContent || textContent.type !== 'text') {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n        inputTokens: response.usage?.input_tokens,\n        outputTokens: response.usage?.output_tokens,\n      });\n      this.logger?.error('No text content in inference response', {\n        model: this.modelId,\n        contentTypes: response.content.map(c => c.type)\n      });\n      throw new Error('No text content in inference response');\n    }\n\n    recordInferenceUsage({\n      provider: this.type,\n      model: this.modelId,\n      durationMs: performance.now() - start,\n      outcome: 'success',\n      inputTokens: response.usage?.input_tokens,\n      outputTokens: response.usage?.output_tokens,\n    });\n\n    // Re-attach the prefill prefix so the returned text is what the caller\n    // *would have seen* if Anthropic had a native JSON mode — a complete,\n    // parseable JSON document. Without this, the consumer would get the\n    // tail (`{...}, {...}]`) and parse-fail trying to read it as a top-\n    // level array.\n    const text = prefill ? prefill + textContent.text : textContent.text;\n\n    this.logger?.info('Text generation completed', {\n      model: this.modelId,\n      textLength: text.length,\n      stopReason: response.stop_reason\n    });\n\n    return {\n      text,\n      stopReason: response.stop_reason || 'unknown'\n    };\n  }\n}\n","// Ollama implementation of InferenceClient interface\n// Uses native Ollama HTTP API (no SDK dependency)\n\nimport type { Logger } from '@semiont/core';\nimport { recordInferenceUsage } from '@semiont/observability';\nimport { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';\n\ninterface OllamaGenerateResponse {\n  response: string;\n  done: boolean;\n  done_reason?: string;\n  /** Number of prompt tokens evaluated. Available on most Ollama versions. */\n  prompt_eval_count?: number;\n  /** Number of tokens generated. */\n  eval_count?: number;\n}\n\nexport class OllamaInferenceClient implements InferenceClient {\n  readonly type = 'ollama' as const;\n  readonly modelId: string;\n  private baseURL: string;\n  private logger?: Logger;\n\n  constructor(model: string, baseURL?: string, logger?: Logger) {\n    this.baseURL = (baseURL || 'http://localhost:11434').replace(/\\/+$/, '');\n    this.modelId = model;\n    this.logger = logger;\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse> {\n    this.logger?.debug('Generating text with Ollama', {\n      model: this.modelId,\n      promptLength: prompt.length,\n      maxTokens,\n      temperature,\n      format: options?.format,\n    });\n\n    const url = `${this.baseURL}/api/generate`;\n    const start = performance.now();\n\n    // Ollama's `format` parameter accepts either the literal string\n    // `\"json\"` (any valid JSON, including objects, numbers, etc.) or a\n    // JSON schema (constrains the top-level shape). The contract on the\n    // inference side is \"parseable JSON array,\" so we pass a minimal\n    // array schema rather than the bare `\"json\"` string — without it,\n    // the model can satisfy \"valid JSON\" by emitting `{\"entities\": [...]}`\n    // and break every consumer that expects to call `.map` on the\n    // top-level value. The schema's `items: {}` keeps element shape\n    // unconstrained — the prompt still carries the per-element schema;\n    // we only enforce the outer array.\n    const body: Record<string, unknown> = {\n      model: this.modelId,\n      prompt,\n      stream: false,\n      think: false,\n      options: {\n        num_predict: maxTokens,\n        temperature,\n      },\n    };\n    if (options?.format === 'json') {\n      body['format'] = { type: 'array', items: {} };\n    }\n\n    let res: Response;\n    try {\n      res = await fetch(url, {\n        method: 'POST',\n        headers: { 'Content-Type': 'application/json' },\n        body: JSON.stringify(body),\n      });\n    } catch (err) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      throw err;\n    }\n\n    if (!res.ok) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n      });\n      const body = await res.text();\n      this.logger?.error('Ollama API error', {\n        model: this.modelId,\n        status: res.status,\n        body,\n      });\n      throw new Error(`Ollama API error (${res.status}): ${body}`);\n    }\n\n    const data = await res.json() as OllamaGenerateResponse;\n\n    if (!data.response) {\n      recordInferenceUsage({\n        provider: this.type,\n        model: this.modelId,\n        durationMs: performance.now() - start,\n        outcome: 'error',\n        inputTokens: data.prompt_eval_count,\n        outputTokens: data.eval_count,\n      });\n      this.logger?.error('Empty response from Ollama', { model: this.modelId });\n      throw new Error('Empty response from Ollama');\n    }\n\n    recordInferenceUsage({\n      provider: this.type,\n      model: this.modelId,\n      durationMs: performance.now() - start,\n      outcome: 'success',\n      inputTokens: data.prompt_eval_count,\n      outputTokens: data.eval_count,\n    });\n\n    const stopReason = mapStopReason(data.done_reason);\n\n    this.logger?.info('Text generation completed', {\n      model: this.modelId,\n      textLength: data.response.length,\n      stopReason,\n    });\n\n    return {\n      text: data.response,\n      stopReason,\n    };\n  }\n}\n\nfunction mapStopReason(doneReason: string | undefined): string {\n  switch (doneReason) {\n    case 'stop': return 'end_turn';\n    case 'length': return 'max_tokens';\n    default: return doneReason || 'unknown';\n  }\n}\n","// Factory for creating inference client instances based on configuration\n\nimport type { Logger } from '@semiont/core';\nimport { InferenceClient } from './interface.js';\nimport { AnthropicInferenceClient } from './implementations/anthropic.js';\nimport { OllamaInferenceClient } from './implementations/ollama.js';\n\nexport type InferenceClientType = 'anthropic' | 'ollama';\n\nexport interface InferenceClientConfig {\n  type: InferenceClientType;\n  apiKey?: string;\n  model: string;\n  endpoint?: string;\n  baseURL?: string;\n}\n\nexport function createInferenceClient(config: InferenceClientConfig, logger?: Logger): InferenceClient {\n  switch (config.type) {\n    case 'anthropic': {\n      if (!config.apiKey || config.apiKey.trim() === '') {\n        throw new Error('apiKey is required for Anthropic inference client');\n      }\n      return new AnthropicInferenceClient(\n        config.apiKey,\n        config.model,\n        config.endpoint || config.baseURL,\n        logger\n      );\n    }\n\n    case 'ollama': {\n      return new OllamaInferenceClient(\n        config.model,\n        config.endpoint || config.baseURL,\n        logger\n      );\n    }\n\n    default:\n      throw new Error(`Unsupported inference client type: ${config.type}`);\n  }\n}\n","// Mock implementation of InferenceClient for testing\n\nimport { InferenceClient, InferenceOptions, InferenceResponse } from '../interface.js';\n\nexport class MockInferenceClient implements InferenceClient {\n  readonly type = 'mock' as const;\n  readonly modelId = 'mock-model' as const;\n  private responses: string[] = [];\n  private responseIndex: number = 0;\n  private stopReasons: string[] = [];\n  public calls: Array<{ prompt: string; maxTokens: number; temperature: number; options?: InferenceOptions }> = [];\n\n  constructor(responses: string[] = ['Mock response'], stopReasons?: string[]) {\n    this.responses = responses;\n    this.stopReasons = stopReasons || responses.map(() => 'end_turn');\n  }\n\n  async generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string> {\n    const response = await this.generateTextWithMetadata(prompt, maxTokens, temperature, options);\n    return response.text;\n  }\n\n  async generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse> {\n    this.calls.push({ prompt, maxTokens, temperature, ...(options ? { options } : {}) });\n\n    const text = this.responses[this.responseIndex];\n    const stopReason = this.stopReasons[this.responseIndex] || 'end_turn';\n\n    if (this.responseIndex < this.responses.length - 1) {\n      this.responseIndex++;\n    }\n\n    return { text, stopReason };\n  }\n\n  // Test helper methods\n  reset(): void {\n    this.calls = [];\n    this.responseIndex = 0;\n  }\n\n  setResponses(responses: string[], stopReasons?: string[]): void {\n    this.responses = responses;\n    this.stopReasons = stopReasons || responses.map(() => 'end_turn');\n    this.responseIndex = 0;\n  }\n}\n"],"mappings":";AAEA,OAAO,eAAe;AAEtB,SAAS,4BAA4B;AAG9B,IAAM,2BAAN,MAA0D;AAAA,EACtD,OAAO;AAAA,EACP;AAAA,EACD;AAAA,EACA;AAAA,EAER,YAAY,QAAgB,OAAe,SAAkB,QAAiB;AAC5E,SAAK,SAAS,IAAI,UAAU;AAAA,MAC1B;AAAA,MACA,SAAS,WAAW;AAAA,IACtB,CAAC;AACD,SAAK,UAAU;AACf,SAAK,SAAS;AAAA,EAChB;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAqB,SAA6C;AACtH,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,aAAa,OAAO;AAC5F,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAqB,SAAwD;AAa7I,UAAM,WAAW,SAAS,WAAW;AACrC,UAAM,UAAU,WAAW,MAAM;AAEjC,SAAK,QAAQ,MAAM,yCAAyC;AAAA,MAC1D,OAAO,KAAK;AAAA,MACZ,cAAc,OAAO;AAAA,MACrB;AAAA,MACA;AAAA,MACA,QAAQ,SAAS;AAAA,IACnB,CAAC;AAED,UAAM,WAAmE;AAAA,MACvE,EAAE,MAAM,QAAQ,SAAS,OAAO;AAAA,IAClC;AACA,QAAI,SAAS;AACX,eAAS,KAAK,EAAE,MAAM,aAAa,SAAS,QAAQ,CAAC;AAAA,IACvD;AAEA,UAAM,QAAQ,YAAY,IAAI;AAC9B,QAAI;AACJ,QAAI;AACF,iBAAW,MAAM,KAAK,OAAO,SAAS,OAAO;AAAA,QAC3C,OAAO,KAAK;AAAA,QACZ,YAAY;AAAA,QACZ;AAAA,QACA;AAAA,MACF,CAAC;AAAA,IACH,SAAS,KAAK;AACZ,2BAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAM;AAAA,IACR;AAEA,SAAK,QAAQ,MAAM,+BAA+B;AAAA,MAChD,OAAO,KAAK;AAAA,MACZ,eAAe,SAAS,QAAQ;AAAA,MAChC,YAAY,SAAS;AAAA,IACvB,CAAC;AAED,UAAM,cAAc,SAAS,QAAQ,KAAK,OAAK,EAAE,SAAS,MAAM;AAEhE,QAAI,CAAC,eAAe,YAAY,SAAS,QAAQ;AAC/C,2BAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,QACT,aAAa,SAAS,OAAO;AAAA,QAC7B,cAAc,SAAS,OAAO;AAAA,MAChC,CAAC;AACD,WAAK,QAAQ,MAAM,yCAAyC;AAAA,QAC1D,OAAO,KAAK;AAAA,QACZ,cAAc,SAAS,QAAQ,IAAI,OAAK,EAAE,IAAI;AAAA,MAChD,CAAC;AACD,YAAM,IAAI,MAAM,uCAAuC;AAAA,IACzD;AAEA,yBAAqB;AAAA,MACnB,UAAU,KAAK;AAAA,MACf,OAAO,KAAK;AAAA,MACZ,YAAY,YAAY,IAAI,IAAI;AAAA,MAChC,SAAS;AAAA,MACT,aAAa,SAAS,OAAO;AAAA,MAC7B,cAAc,SAAS,OAAO;AAAA,IAChC,CAAC;AAOD,UAAM,OAAO,UAAU,UAAU,YAAY,OAAO,YAAY;AAEhE,SAAK,QAAQ,KAAK,6BAA6B;AAAA,MAC7C,OAAO,KAAK;AAAA,MACZ,YAAY,KAAK;AAAA,MACjB,YAAY,SAAS;AAAA,IACvB,CAAC;AAED,WAAO;AAAA,MACL;AAAA,MACA,YAAY,SAAS,eAAe;AAAA,IACtC;AAAA,EACF;AACF;;;AC5HA,SAAS,wBAAAA,6BAA4B;AAa9B,IAAM,wBAAN,MAAuD;AAAA,EACnD,OAAO;AAAA,EACP;AAAA,EACD;AAAA,EACA;AAAA,EAER,YAAY,OAAe,SAAkB,QAAiB;AAC5D,SAAK,WAAW,WAAW,0BAA0B,QAAQ,QAAQ,EAAE;AACvE,SAAK,UAAU;AACf,SAAK,SAAS;AAAA,EAChB;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAqB,SAA6C;AACtH,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,aAAa,OAAO;AAC5F,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAqB,SAAwD;AAC7I,SAAK,QAAQ,MAAM,+BAA+B;AAAA,MAChD,OAAO,KAAK;AAAA,MACZ,cAAc,OAAO;AAAA,MACrB;AAAA,MACA;AAAA,MACA,QAAQ,SAAS;AAAA,IACnB,CAAC;AAED,UAAM,MAAM,GAAG,KAAK,OAAO;AAC3B,UAAM,QAAQ,YAAY,IAAI;AAY9B,UAAM,OAAgC;AAAA,MACpC,OAAO,KAAK;AAAA,MACZ;AAAA,MACA,QAAQ;AAAA,MACR,OAAO;AAAA,MACP,SAAS;AAAA,QACP,aAAa;AAAA,QACb;AAAA,MACF;AAAA,IACF;AACA,QAAI,SAAS,WAAW,QAAQ;AAC9B,WAAK,QAAQ,IAAI,EAAE,MAAM,SAAS,OAAO,CAAC,EAAE;AAAA,IAC9C;AAEA,QAAI;AACJ,QAAI;AACF,YAAM,MAAM,MAAM,KAAK;AAAA,QACrB,QAAQ;AAAA,QACR,SAAS,EAAE,gBAAgB,mBAAmB;AAAA,QAC9C,MAAM,KAAK,UAAU,IAAI;AAAA,MAC3B,CAAC;AAAA,IACH,SAAS,KAAK;AACZ,MAAAA,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAM;AAAA,IACR;AAEA,QAAI,CAAC,IAAI,IAAI;AACX,MAAAA,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,MACX,CAAC;AACD,YAAMC,QAAO,MAAM,IAAI,KAAK;AAC5B,WAAK,QAAQ,MAAM,oBAAoB;AAAA,QACrC,OAAO,KAAK;AAAA,QACZ,QAAQ,IAAI;AAAA,QACZ,MAAAA;AAAA,MACF,CAAC;AACD,YAAM,IAAI,MAAM,qBAAqB,IAAI,MAAM,MAAMA,KAAI,EAAE;AAAA,IAC7D;AAEA,UAAM,OAAO,MAAM,IAAI,KAAK;AAE5B,QAAI,CAAC,KAAK,UAAU;AAClB,MAAAD,sBAAqB;AAAA,QACnB,UAAU,KAAK;AAAA,QACf,OAAO,KAAK;AAAA,QACZ,YAAY,YAAY,IAAI,IAAI;AAAA,QAChC,SAAS;AAAA,QACT,aAAa,KAAK;AAAA,QAClB,cAAc,KAAK;AAAA,MACrB,CAAC;AACD,WAAK,QAAQ,MAAM,8BAA8B,EAAE,OAAO,KAAK,QAAQ,CAAC;AACxE,YAAM,IAAI,MAAM,4BAA4B;AAAA,IAC9C;AAEA,IAAAA,sBAAqB;AAAA,MACnB,UAAU,KAAK;AAAA,MACf,OAAO,KAAK;AAAA,MACZ,YAAY,YAAY,IAAI,IAAI;AAAA,MAChC,SAAS;AAAA,MACT,aAAa,KAAK;AAAA,MAClB,cAAc,KAAK;AAAA,IACrB,CAAC;AAED,UAAM,aAAa,cAAc,KAAK,WAAW;AAEjD,SAAK,QAAQ,KAAK,6BAA6B;AAAA,MAC7C,OAAO,KAAK;AAAA,MACZ,YAAY,KAAK,SAAS;AAAA,MAC1B;AAAA,IACF,CAAC;AAED,WAAO;AAAA,MACL,MAAM,KAAK;AAAA,MACX;AAAA,IACF;AAAA,EACF;AACF;AAEA,SAAS,cAAc,YAAwC;AAC7D,UAAQ,YAAY;AAAA,IAClB,KAAK;AAAQ,aAAO;AAAA,IACpB,KAAK;AAAU,aAAO;AAAA,IACtB;AAAS,aAAO,cAAc;AAAA,EAChC;AACF;;;ACnIO,SAAS,sBAAsB,QAA+B,QAAkC;AACrG,UAAQ,OAAO,MAAM;AAAA,IACnB,KAAK,aAAa;AAChB,UAAI,CAAC,OAAO,UAAU,OAAO,OAAO,KAAK,MAAM,IAAI;AACjD,cAAM,IAAI,MAAM,mDAAmD;AAAA,MACrE;AACA,aAAO,IAAI;AAAA,QACT,OAAO;AAAA,QACP,OAAO;AAAA,QACP,OAAO,YAAY,OAAO;AAAA,QAC1B;AAAA,MACF;AAAA,IACF;AAAA,IAEA,KAAK,UAAU;AACb,aAAO,IAAI;AAAA,QACT,OAAO;AAAA,QACP,OAAO,YAAY,OAAO;AAAA,QAC1B;AAAA,MACF;AAAA,IACF;AAAA,IAEA;AACE,YAAM,IAAI,MAAM,sCAAsC,OAAO,IAAI,EAAE;AAAA,EACvE;AACF;;;ACtCO,IAAM,sBAAN,MAAqD;AAAA,EACjD,OAAO;AAAA,EACP,UAAU;AAAA,EACX,YAAsB,CAAC;AAAA,EACvB,gBAAwB;AAAA,EACxB,cAAwB,CAAC;AAAA,EAC1B,QAAuG,CAAC;AAAA,EAE/G,YAAY,YAAsB,CAAC,eAAe,GAAG,aAAwB;AAC3E,SAAK,YAAY;AACjB,SAAK,cAAc,eAAe,UAAU,IAAI,MAAM,UAAU;AAAA,EAClE;AAAA,EAEA,MAAM,aAAa,QAAgB,WAAmB,aAAqB,SAA6C;AACtH,UAAM,WAAW,MAAM,KAAK,yBAAyB,QAAQ,WAAW,aAAa,OAAO;AAC5F,WAAO,SAAS;AAAA,EAClB;AAAA,EAEA,MAAM,yBAAyB,QAAgB,WAAmB,aAAqB,SAAwD;AAC7I,SAAK,MAAM,KAAK,EAAE,QAAQ,WAAW,aAAa,GAAI,UAAU,EAAE,QAAQ,IAAI,CAAC,EAAG,CAAC;AAEnF,UAAM,OAAO,KAAK,UAAU,KAAK,aAAa;AAC9C,UAAM,aAAa,KAAK,YAAY,KAAK,aAAa,KAAK;AAE3D,QAAI,KAAK,gBAAgB,KAAK,UAAU,SAAS,GAAG;AAClD,WAAK;AAAA,IACP;AAEA,WAAO,EAAE,MAAM,WAAW;AAAA,EAC5B;AAAA;AAAA,EAGA,QAAc;AACZ,SAAK,QAAQ,CAAC;AACd,SAAK,gBAAgB;AAAA,EACvB;AAAA,EAEA,aAAa,WAAqB,aAA8B;AAC9D,SAAK,YAAY;AACjB,SAAK,cAAc,eAAe,UAAU,IAAI,MAAM,UAAU;AAChE,SAAK,gBAAgB;AAAA,EACvB;AACF;","names":["recordInferenceUsage","body"]}

package/dist/interface.d.ts

@@ -0,0 +1,41 @@
+export interface InferenceResponse {
+    text: string;
+    stopReason: 'end_turn' | 'max_tokens' | 'stop_sequence' | string;
+}
+/**
+ * Per-call options. Drift here when output discipline matters more than
+ * raw model behavior — e.g. forcing valid JSON for entity extraction
+ * where a parse failure in the consumer is silent and useless.
+ */
+export interface InferenceOptions {
+    /**
+     * Constrain output to a parseable JSON array. Every implementation
+     * MUST satisfy this contract using whatever mechanism its provider
+     * supports — Ollama uses grammar-constrained sampling
+     * (`format: "json"`); Anthropic uses assistant-turn prefill (`[`)
+     * and re-attaches the prefix on return. Callers can rely on the
+     * returned `text` being a top-level JSON array regardless of
+     * provider.
+     *
+     * Current callers all expect arrays (entity extraction, motivation
+     * detection). If an object-emitting caller appears, this option
+     * grows a `root: 'array' | 'object'` field; do not silently drop
+     * the constraint.
+     */
+    format?: 'json';
+}
+export interface InferenceClient {
+    /** Provider type identifier (e.g. 'anthropic', 'ollama') */
+    readonly type: string;
+    /** Model identifier used for generation (e.g. 'claude-opus-4-6', 'llama3') */
+    readonly modelId: string;
+    /**
+     * Generate text from a prompt (simple interface)
+     */
+    generateText(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<string>;
+    /**
+     * Generate text with detailed response information
+     */
+    generateTextWithMetadata(prompt: string, maxTokens: number, temperature: number, options?: InferenceOptions): Promise<InferenceResponse>;
+}
+//# sourceMappingURL=interface.d.ts.map

package/dist/interface.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"interface.d.ts","sourceRoot":"","sources":["../src/interface.ts"],"names":[],"mappings":"AAEA,MAAM,WAAW,iBAAiB;IAChC,IAAI,EAAE,MAAM,CAAC;IACb,UAAU,EAAE,UAAU,GAAG,YAAY,GAAG,eAAe,GAAG,MAAM,CAAC;CAClE;AAED;;;;GAIG;AACH,MAAM,WAAW,gBAAgB;IAC/B;;;;;;;;;;;;;OAaG;IACH,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,eAAe;IAC9B,4DAA4D;IAC5D,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IAEtB,8EAA8E;IAC9E,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IAEzB;;OAEG;IACH,YAAY,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC;IAElH;;OAEG;IACH,wBAAwB,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,WAAW,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,gBAAgB,GAAG,OAAO,CAAC,iBAAiB,CAAC,CAAC;CAC1I"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@semiont/inference",
-  "version": "0.5.2",
+  "version": "0.5.4",
   "type": "module",
   "description": "AI inference capabilities for entity extraction, text generation, and resource creation",
   "main": "./dist/index.js",
@@ -16,7 +16,7 @@
     "README.md"
   ],
   "scripts": {
-    "build": "npm run typecheck && tsup",
+    "build": "npm run typecheck && tsup && tsc -p tsconfig.build.json",
     "typecheck": "tsc --noEmit",
     "clean": "rm -rf dist",
     "test": "vitest run",
@@ -32,7 +32,7 @@
   "devDependencies": {
     "@vitest/coverage-v8": "^4.1.0",
     "tsup": "^8.0.1",
-    "typescript": "^5.6.3"
+    "typescript": "^6.0.2"
   },
   "publishConfig": {
     "access": "public"