Add Ollama provider for self-hosted LLM inference

Adds LLM_PROVIDER=ollama for fully local, zero-cost inference
via Ollama's OpenAI-compatible API. No API key required.
Configurable base URL via OLLAMA_BASE_URL env var.

test/llm-ollama-integration.test.mjs

@@ -0,0 +1,36 @@
+// Ollama provider — integration test (calls real Ollama instance)
+// Requires a running Ollama server with a model pulled
+// Run: OLLAMA_MODEL=llama3.1:8b node --test test/llm-ollama-integration.test.mjs
+
+import { describe, it } from 'node:test';
+import assert from 'node:assert/strict';
+import { OllamaProvider } from '../lib/llm/ollama.mjs';
+
+const BASE_URL = process.env.OLLAMA_BASE_URL || 'http://localhost:11434';
+const MODEL = process.env.OLLAMA_MODEL || 'llama3.1:8b';
+
+// Check if Ollama is reachable before running tests
+let ollamaAvailable = false;
+try {
+  const res = await fetch(`${BASE_URL}/api/tags`, { signal: AbortSignal.timeout(3000) });
+  ollamaAvailable = res.ok;
+} catch { /* not available */ }
+
+describe('Ollama integration', { skip: !ollamaAvailable && 'Ollama not reachable' }, () => {
+  it('should complete a prompt via local Ollama', async () => {
+    const provider = new OllamaProvider({ model: MODEL, baseUrl: BASE_URL });
+    assert.equal(provider.isConfigured, true);
+
+    const result = await provider.complete(
+      'You are a helpful assistant. Respond in exactly one sentence.',
+      'What is 2+2?',
+      { maxTokens: 128, timeout: 60000 }
+    );
+
+    assert.ok(result.text.length > 0, 'Response text should not be empty');
+    assert.ok(result.model, 'Should report model name');
+    console.log(`  Response: ${result.text}`);
+    console.log(`  Tokens: ${result.usage.inputTokens} in / ${result.usage.outputTokens} out`);
+    console.log(`  Model: ${result.model}`);
+  });
+});