feat: add heartbeat monitor and vector memory search (Tier 2)

Heartbeat:
- HeartbeatMonitor with 5 checks: gateway, model, channels, memory, disk
- Configurable interval, failure threshold, notification channel
- Recovery notifications when health restores
- 25 new tests

Vector Memory Search:
- EmbeddingProvider interface with OpenAI, Gemini, Ollama, LlamaCpp backends
- SQLite-backed VectorStore with cosine similarity search
- Text chunker with paragraph-aware splitting and overlap
- HybridSearch merging keyword + vector results with configurable weight
- Background indexer with dirty-namespace tracking
- Graceful fallback to keyword search when embeddings unavailable
- 51 new tests

Config: automation.heartbeat + memory.embedding schema sections
Total: 950 tests passing, all types clean

src/memory/embeddings.ts

@@ -0,0 +1,182 @@
+/**
+ * Embedding provider interface and implementations for multiple backends.
+ */
+
+import type { EmbeddingConfig } from '../config/schema.js';
+
+/**
+ * Interface for embedding providers that convert text to vectors.
+ */
+export interface EmbeddingProvider {
+  /** Generate embeddings for one or more texts. Returns one vector per text. */
+  embed(texts: string[]): Promise<number[][]>;
+  /** The dimensionality of the embedding vectors. */
+  dimensions: number;
+}
+
+// ---------------------------------------------------------------------------
+// OpenAI
+// ---------------------------------------------------------------------------
+
+export class OpenAIEmbeddingProvider implements EmbeddingProvider {
+  private _model: string;
+  private _dimensions: number;
+  private _apiKey: string;
+  private _endpoint?: string;
+
+  constructor(config: EmbeddingConfig) {
+    this._model = config.model;
+    this._dimensions = config.dimensions ?? 1536;
+    this._apiKey = config.api_key ?? process.env.OPENAI_API_KEY ?? '';
+    this._endpoint = config.endpoint;
+  }
+
+  get dimensions(): number {
+    return this._dimensions;
+  }
+
+  async embed(texts: string[]): Promise<number[][]> {
+    const { default: OpenAI } = await import('openai');
+    const client = new OpenAI({
+      apiKey: this._apiKey,
+      ...(this._endpoint ? { baseURL: this._endpoint } : {}),
+    });
+
+    const response = await client.embeddings.create({
+      model: this._model,
+      input: texts,
+      ...(this._dimensions ? { dimensions: this._dimensions } : {}),
+    });
+
+    // Sort by index to ensure order matches input
+    const sorted = response.data.sort((a, b) => a.index - b.index);
+    return sorted.map((item) => item.embedding);
+  }
+}
+
+// ---------------------------------------------------------------------------
+// Gemini
+// ---------------------------------------------------------------------------
+
+export class GeminiEmbeddingProvider implements EmbeddingProvider {
+  private _model: string;
+  private _dimensions: number;
+  private _apiKey: string;
+
+  constructor(config: EmbeddingConfig) {
+    this._model = config.model;
+    this._dimensions = config.dimensions ?? 768;
+    this._apiKey = config.api_key ?? process.env.GOOGLE_API_KEY ?? '';
+  }
+
+  get dimensions(): number {
+    return this._dimensions;
+  }
+
+  async embed(texts: string[]): Promise<number[][]> {
+    const { GoogleGenerativeAI } = await import('@google/generative-ai');
+    const genAI = new GoogleGenerativeAI(this._apiKey);
+    const model = genAI.getGenerativeModel({ model: this._model });
+
+    // Use batchEmbedContents for efficiency
+    const requests = texts.map((text) => ({
+      content: { role: 'user' as const, parts: [{ text }] },
+    }));
+
+    const response = await model.batchEmbedContents({ requests });
+    return response.embeddings.map((e) => e.values);
+  }
+}
+
+// ---------------------------------------------------------------------------
+// Ollama
+// ---------------------------------------------------------------------------
+
+export class OllamaEmbeddingProvider implements EmbeddingProvider {
+  private _model: string;
+  private _dimensions: number;
+  private _host?: string;
+
+  constructor(config: EmbeddingConfig) {
+    this._model = config.model;
+    this._dimensions = config.dimensions ?? 768;
+    this._host = config.endpoint;
+  }
+
+  get dimensions(): number {
+    return this._dimensions;
+  }
+
+  async embed(texts: string[]): Promise<number[][]> {
+    const { Ollama } = await import('ollama');
+    const client = new Ollama({ host: this._host });
+
+    const response = await client.embed({
+      model: this._model,
+      input: texts,
+    });
+
+    return response.embeddings;
+  }
+}
+
+// ---------------------------------------------------------------------------
+// LlamaCpp
+// ---------------------------------------------------------------------------
+
+export class LlamaCppEmbeddingProvider implements EmbeddingProvider {
+  private _dimensions: number;
+  private _endpoint: string;
+
+  constructor(config: EmbeddingConfig) {
+    this._dimensions = config.dimensions ?? 768;
+    this._endpoint = config.endpoint ?? 'http://localhost:8080';
+  }
+
+  get dimensions(): number {
+    return this._dimensions;
+  }
+
+  async embed(texts: string[]): Promise<number[][]> {
+    const results: number[][] = [];
+
+    for (const text of texts) {
+      const response = await fetch(`${this._endpoint}/embedding`, {
+        method: 'POST',
+        headers: { 'Content-Type': 'application/json' },
+        body: JSON.stringify({ content: text }),
+      });
+
+      if (!response.ok) {
+        throw new Error(`LlamaCpp embedding request failed: ${response.status} ${response.statusText}`);
+      }
+
+      const data = (await response.json()) as { embedding: number[] };
+      results.push(data.embedding);
+    }
+
+    return results;
+  }
+}
+
+// ---------------------------------------------------------------------------
+// Factory
+// ---------------------------------------------------------------------------
+
+/**
+ * Create an embedding provider from config.
+ */
+export function createEmbeddingProvider(config: EmbeddingConfig): EmbeddingProvider {
+  switch (config.provider) {
+    case 'openai':
+      return new OpenAIEmbeddingProvider(config);
+    case 'gemini':
+      return new GeminiEmbeddingProvider(config);
+    case 'ollama':
+      return new OllamaEmbeddingProvider(config);
+    case 'llamacpp':
+      return new LlamaCppEmbeddingProvider(config);
+    default:
+      throw new Error(`Unknown embedding provider: ${(config as Record<string, unknown>).provider}`);
+  }
+}