universal-llm-client 4.3.0 → 4.5.0

package/src/tests/providers/anthropic.test.ts

@@ -0,0 +1,118 @@
+/**
+ * Anthropic Provider — unified thinking flag tests.
+ *
+ * Validates that the model-level `thinking` config and per-call
+ * `ChatOptions.thinking` map to Anthropic extended thinking
+ * (`thinking: { type: 'enabled', budget_tokens }`), with the API's
+ * constraints handled (budget < max_tokens; temperature omitted when on).
+ */
+
+import { describe, test, expect, mock, beforeEach, afterEach } from 'bun:test';
+import { AnthropicClient } from '../../providers/anthropic.js';
+import type { LLMClientOptions } from '../../interfaces.js';
+import { AIModelApiType } from '../../interfaces.js';
+
+function createClient(overrides?: Partial<LLMClientOptions>): AnthropicClient {
+    return new AnthropicClient({
+        model: 'claude-sonnet-4-5',
+        apiKey: 'test-api-key',
+        apiType: AIModelApiType.Anthropic,
+        ...overrides,
+    });
+}
+
+const ANTHROPIC_RESPONSE = {
+    id: 'msg_1',
+    type: 'message',
+    role: 'assistant',
+    content: [{ type: 'text', text: 'Hi there.' }],
+    model: 'claude-sonnet-4-5',
+    stop_reason: 'end_turn',
+    usage: { input_tokens: 5, output_tokens: 3 },
+};
+
+describe('AnthropicClient thinking flag', () => {
+    let originalFetch: typeof globalThis.fetch;
+    beforeEach(() => { originalFetch = globalThis.fetch; });
+    afterEach(() => { globalThis.fetch = originalFetch; });
+
+    function mockFetchAndCapture(response: unknown = ANTHROPIC_RESPONSE, status = 200) {
+        let capturedBody: Record<string, unknown> | null = null;
+        globalThis.fetch = mock(async (_input: string | URL | Request, init?: RequestInit) => {
+            if (init?.body) capturedBody = JSON.parse(init.body as string);
+            return new Response(JSON.stringify(response), {
+                status,
+                headers: { 'Content-Type': 'application/json' },
+            });
+        }) as typeof fetch;
+        return () => capturedBody;
+    }
+
+    test('enables extended thinking when thinking:true (budget < max_tokens, temperature omitted)', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient();
+
+        await client.chat([{ role: 'user', content: 'hi' }], { thinking: true, maxTokens: 4096, temperature: 0.7 });
+
+        const body = getBody()!;
+        const thinking = body['thinking'] as Record<string, unknown> | undefined;
+        expect(thinking).toBeDefined();
+        expect(thinking!['type']).toBe('enabled');
+        expect(thinking!['budget_tokens']).toBe(2048);
+        expect(thinking!['budget_tokens'] as number).toBeLessThan(body['max_tokens'] as number);
+        // The API forbids a custom temperature while thinking is on — must be omitted.
+        expect(body['temperature']).toBeUndefined();
+    });
+
+    test('does not request thinking when flag is unset (temperature preserved)', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient();
+
+        await client.chat([{ role: 'user', content: 'hi' }], { temperature: 0.5 });
+
+        const body = getBody()!;
+        expect(body['thinking']).toBeUndefined();
+        expect(body['temperature']).toBe(0.5);
+    });
+
+    test('does not request thinking when thinking:false (overrides client config)', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient({ thinking: true });
+
+        await client.chat([{ role: 'user', content: 'hi' }], { thinking: false });
+
+        expect(getBody()!['thinking']).toBeUndefined();
+    });
+
+    test('per-call thinking:true overrides unset client config', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient();
+
+        await client.chat([{ role: 'user', content: 'hi' }], { thinking: true });
+
+        expect((getBody()!['thinking'] as Record<string, unknown>)['type']).toBe('enabled');
+    });
+
+    test('maps a thinking level to budget_tokens (high, clamped < max_tokens)', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient();
+
+        await client.chat([{ role: 'user', content: 'hi' }], { thinking: 'high', maxTokens: 32000 });
+
+        const thinking = getBody()!['thinking'] as Record<string, unknown>;
+        expect(thinking['type']).toBe('enabled');
+        expect(thinking['budget_tokens']).toBe(16384);
+    });
+
+    test('bumps max_tokens so budget_tokens stays below it when maxTokens is small', async () => {
+        const getBody = mockFetchAndCapture();
+        const client = createClient();
+
+        await client.chat([{ role: 'user', content: 'hi' }], { thinking: 'low', maxTokens: 1024 });
+
+        const body = getBody()!;
+        const budget = (body['thinking'] as Record<string, unknown>)['budget_tokens'] as number;
+        expect(budget).toBeGreaterThanOrEqual(1024);
+        expect(budget).toBeLessThan(body['max_tokens'] as number);
+    });
+});