badgr-cli 1.0.35 → 1.0.36

package/tests/serve-lifecycle.test.js

@@ -22,6 +22,7 @@ import { serveCommand } from '../src/commands/serve.js';
 vi.mock('../src/api.js', () => ({
   callApi:             vi.fn(),
   terminateDeployment: vi.fn().mockResolvedValue({}),
+  listDeployments:     vi.fn().mockResolvedValue({ deployments: [], count: 0 }),
 }));
 
 vi.mock('../src/store.js', () => ({
@@ -107,7 +108,7 @@ describe('successful serve', () => {
       text: async () => '',
     });
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--gpu', 'L40S'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--gpu', 'L40S', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(5000);
     await p;
 
@@ -131,7 +132,7 @@ describe('successful serve', () => {
 
     global.fetch = vi.fn().mockResolvedValue({ ok: true, status: 200, json: async () => ({}) });
 
-    const p = serveCommand(config, ['mistralai/Mistral-7B-v0.1', '--gpu', 'A100'], chalk);
+    const p = serveCommand(config, ['mistralai/Mistral-7B-v0.1', '--gpu', 'A100', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(5000);
     await p;
 
@@ -156,7 +157,7 @@ describe('billing ends when deployment stops', () => {
 
     global.fetch = vi.fn().mockResolvedValue({ ok: true, status: 200, json: async () => ({}) });
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(5000);
     await p;
 
@@ -182,7 +183,7 @@ describe('deployment fails during startup', () => {
       .mockResolvedValueOnce(makeServeDep()) // POST /serve
       .mockResolvedValueOnce({ status: 'failed', error: 'OOM: not enough VRAM' }); // pre-health check
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(2000);
     await p;
 
@@ -201,7 +202,7 @@ describe('deployment fails during startup', () => {
     // Health check endpoint doesn't respond (dep crashed)
     global.fetch = vi.fn().mockRejectedValue(new Error('Connection refused'));
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
     // need to advance past the 8000ms sleep in waitForEndpoint loop
     await vi.advanceTimersByTimeAsync(10000);
     await p;
@@ -227,7 +228,7 @@ describe('health check timeout', () => {
     // Health endpoint never responds
     global.fetch = vi.fn().mockRejectedValue(new Error('ETIMEDOUT'));
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
     // waitForEndpoint timeout is 5 * 60 * 1000 = 300000ms
     await vi.advanceTimersByTimeAsync(310000);
     await p;
@@ -256,7 +257,7 @@ describe('invalid model ID or no capacity', () => {
     const errLines = [];
     console.error.mockImplementation(msg => errLines.push(msg));
 
-    await serveCommand(config, ['invalid/nonexistent-model-xyz'], chalk);
+    await serveCommand(config, ['invalid/nonexistent-model-xyz', '--max-cost', '5'], chalk);
 
     expect(process.exitCode).toBe(1);
     const combined = errLines.join('\n');
@@ -276,7 +277,7 @@ describe('invalid model ID or no capacity', () => {
     const errLines = [];
     console.error.mockImplementation(msg => errLines.push(msg));
 
-    await serveCommand(config, ['my-org/private-model'], chalk);
+    await serveCommand(config, ['my-org/private-model', '--max-cost', '5'], chalk);
 
     expect(process.exitCode).toBe(1);
     // Debug info hidden by default (BADGR_DEBUG not set)
@@ -293,7 +294,7 @@ describe('invalid model ID or no capacity', () => {
     });
     api.callApi.mockRejectedValue(provErr);
 
-    await serveCommand(config, ['some/model', '--task', 'unknown_task'], chalk);
+    await serveCommand(config, ['some/model', '--task', 'unknown_task', '--max-cost', '5'], chalk);
 
     expect(process.exitCode).toBe(1);
     expect(store.addDeployment).not.toHaveBeenCalled();
@@ -309,7 +310,7 @@ describe('--no-wait flag', () => {
     api.callApi.mockResolvedValueOnce(makeServeDep());
     global.fetch = vi.fn(); // should never be called
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--no-wait'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--no-wait', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(100);
     await p;
 
@@ -333,7 +334,7 @@ describe('custom image health check', () => {
 
     const p = serveCommand(
       config,
-      ['--image', 'ghcr.io/my-org/diffusers-api:latest', '--gpu', 'L40S'],
+      ['--image', 'ghcr.io/my-org/diffusers-api:latest', '--gpu', 'L40S', '--max-cost', '5'],
       chalk,
     );
     await vi.advanceTimersByTimeAsync(100);
@@ -356,7 +357,7 @@ describe('custom image health check', () => {
 
     const p = serveCommand(
       config,
-      ['--image', 'ghcr.io/my-org/comfyui-custom:latest', '--gpu', 'A100'],
+      ['--image', 'ghcr.io/my-org/comfyui-custom:latest', '--gpu', 'A100', '--max-cost', '5'],
       chalk,
     );
     await vi.advanceTimersByTimeAsync(5000);
@@ -380,7 +381,7 @@ describe('custom image health check', () => {
 
     const p = serveCommand(
       config,
-      ['--image', 'ghcr.io/my-org/stable-diffusion:latest', '--health-path', '/health'],
+      ['--image', 'ghcr.io/my-org/stable-diffusion:latest', '--health-path', '/health', '--max-cost', '5'],
       chalk,
     );
     await vi.advanceTimersByTimeAsync(5000);
@@ -404,7 +405,7 @@ describe('custom image health check', () => {
 
     const p = serveCommand(
       config,
-      ['meta-llama/Llama-3.1-8B-Instruct', '--health-path', '/v1/health'],
+      ['meta-llama/Llama-3.1-8B-Instruct', '--health-path', '/v1/health', '--max-cost', '5'],
       chalk,
     );
     await vi.advanceTimersByTimeAsync(5000);
@@ -433,7 +434,7 @@ describe('tier-2 fallback for serve', () => {
 
     global.fetch = vi.fn().mockResolvedValue({ ok: true, status: 200, json: async () => ({}) });
 
-    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    const p = serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
     await vi.advanceTimersByTimeAsync(5000);
     await p;
 
@@ -450,7 +451,7 @@ describe('tier-2 fallback for serve', () => {
     });
     api.callApi.mockRejectedValueOnce(capacityErr);
 
-    await serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--no-fallback'], chalk);
+    await serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--no-fallback', '--max-cost', '5'], chalk);
 
     const postCalls = api.callApi.mock.calls.filter(c => c[1]?.method === 'POST');
     expect(postCalls.length).toBe(1);
@@ -466,7 +467,7 @@ describe('missing endpoint URL guard', () => {
   it('exits with error when backend returns no endpoint_url', async () => {
     api.callApi.mockResolvedValueOnce(makeServeDep({ endpoint_url: null, openai_base_url: null }));
 
-    await serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct'], chalk);
+    await serveCommand(config, ['meta-llama/Llama-3.1-8B-Instruct', '--max-cost', '5'], chalk);
 
     expect(process.exitCode).toBe(1);
     expect(store.updateReceipt).toHaveBeenCalledWith('rcpt-serve-001', expect.objectContaining({