amalfa 1.0.28 → 1.0.29

package/src/daemon/sonar-inference.ts

@@ -0,0 +1,116 @@
+import { AMALFA_DIRS, loadConfig } from "@src/config/defaults";
+import { getLogger } from "@src/utils/Logger";
+import type { Message, RequestOptions } from "./sonar-types";
+
+const log = getLogger("SonarInference");
+
+/**
+ * Shared state for inference
+ */
+export const inferenceState = {
+	ollamaAvailable: false,
+	ollamaModel: "phi3:latest",
+};
+
+/**
+ * Call Ollama HTTP API for inference
+ * Supports both local Ollama and OpenRouter (cloud) providers.
+ */
+export async function callOllama(
+	messages: Message[],
+	options: RequestOptions = {},
+): Promise<{ message: Message }> {
+	const config = await loadConfig();
+	const hostArgs = config.sonar;
+
+	// Cloud toggle: dev-cloud/prod-local strategy
+	const cloudConfig = hostArgs.cloud;
+	const useCloud = cloudConfig?.enabled === true;
+	const provider = useCloud ? cloudConfig.provider || "ollama" : "ollama";
+
+	// Tiered model strategy: options.model > cloud.model > discovered > config > default
+	const { format, model: overrideModel, ...modelOptions } = options;
+	const model =
+		overrideModel ||
+		(useCloud ? cloudConfig.model : null) ||
+		inferenceState.ollamaModel ||
+		hostArgs.model ||
+		"qwen2.5:1.5b";
+
+	// Build headers
+	const headers: Record<string, string> = {
+		"Content-Type": "application/json",
+	};
+	// API key: prefer env var (OPENROUTER_API_KEY) over config
+	const apiKey = process.env.OPENROUTER_API_KEY || cloudConfig?.apiKey;
+	if (useCloud && apiKey) {
+		headers.Authorization = `Bearer ${apiKey}`;
+		log.info(
+			{ provider, hasKey: !!apiKey, keyLength: apiKey?.length },
+			"Cloud request with API key",
+		);
+	} else if (useCloud) {
+		log.warn("Cloud enabled but no API key found in env or config!");
+	}
+
+	// OpenRouter requires site headers for tracking
+	if (provider === "openrouter") {
+		headers["HTTP-Referer"] = "https://github.com/pjsvis/amalfa";
+		headers["X-Title"] = "AMALFA Knowledge Graph";
+	}
+
+	// Determine endpoint and request format based on provider
+	let endpoint: string;
+	let body: string;
+
+	if (provider === "openrouter") {
+		endpoint = "https://openrouter.ai/api/v1/chat/completions";
+		body = JSON.stringify({
+			model,
+			messages,
+			stream: false,
+			...modelOptions,
+		});
+	} else {
+		// Local Ollama
+		const host = hostArgs.host || "localhost:11434";
+		endpoint = `http://${host}/api/chat`;
+		body = JSON.stringify({
+			model,
+			messages,
+			stream: false,
+			format,
+			options: modelOptions,
+		});
+	}
+
+	try {
+		const response = await fetch(endpoint, {
+			method: "POST",
+			headers,
+			body,
+		});
+
+		if (!response.ok) {
+			const error = await response.text();
+			throw new Error(`Inference failed (${response.status}): ${error}`);
+		}
+
+		const result = await response.json();
+
+		if (provider === "openrouter") {
+			// OpenAI format
+			return {
+				message: (result as any).choices[0].message,
+			};
+		}
+		// Ollama format
+		return {
+			message: (result as any).message,
+		};
+	} catch (error) {
+		const errorMsg = error instanceof Error ? error.message : String(error);
+		log.error({ error: errorMsg, endpoint, model }, "Ollama inference failed");
+		throw error;
+	}
+}