@vellumai/assistant 0.4.56 → 0.4.57

package/src/__tests__/assistant-attachments.test.ts

@@ -193,11 +193,11 @@ describe("validateDrafts", () => {
     expect(result.warnings).toHaveLength(0);
   });
 
-  test("accepts a 25 MB attachment (under 50 MB limit)", () => {
+  test("accepts a 75 MB attachment (under 100 MB limit)", () => {
     const drafts = [
       makeDraft({
         filename: "video.mov",
-        sizeBytes: 25 * 1024 * 1024,
+        sizeBytes: 75 * 1024 * 1024,
       }),
     ];
     const result = validateDrafts(drafts);
@@ -205,18 +205,18 @@ describe("validateDrafts", () => {
     expect(result.warnings).toHaveLength(0);
   });
 
-  test("rejects a 51 MB attachment with warning mentioning 50.0 MB limit", () => {
+  test("rejects a 101 MB attachment with warning mentioning 100.0 MB limit", () => {
     const drafts = [
       makeDraft({
         filename: "huge.mov",
-        sizeBytes: 51 * 1024 * 1024,
+        sizeBytes: 101 * 1024 * 1024,
       }),
     ];
     const result = validateDrafts(drafts);
     expect(result.accepted).toHaveLength(0);
     expect(result.warnings).toHaveLength(1);
     expect(result.warnings[0]).toContain("huge.mov");
-    expect(result.warnings[0]).toContain("50.0 MB");
+    expect(result.warnings[0]).toContain("100.0 MB");
     expect(result.warnings[0]).toContain("limit");
   });
 });

package/src/__tests__/assistant-events-sse-hardening.test.ts

@@ -41,7 +41,7 @@ mock.module("../config/loader.js", () => ({
     model: "test",
     provider: "test",
     memory: { enabled: false },
-    rateLimit: { maxRequestsPerMinute: 0, maxTokensPerSession: 0 },
+    rateLimit: { maxRequestsPerMinute: 0 },
     secretDetection: { enabled: false },
   }),
 }));

package/src/__tests__/assistant-feature-flags-integration.test.ts

@@ -29,7 +29,6 @@ const TEST_DIR = join(
 );
 
 let currentConfig: Record<string, unknown> = {
-  sandbox: { enabled: false, backend: "native" },
   services: {
     inference: {
       mode: "your-own",
@@ -39,9 +38,9 @@ let currentConfig: Record<string, unknown> = {
     "image-generation": {
       mode: "your-own",
       provider: "gemini",
-      model: "gemini-2.5-flash-image",
+      model: "gemini-3.1-flash-image-preview",
     },
-    "web-search": { mode: "your-own", provider: "anthropic-native" },
+    "web-search": { mode: "your-own", provider: "inference-provider-native" },
   },
 };
 
@@ -132,7 +131,6 @@ const { skillFlagKey } = await import("../config/skill-state.js");
 beforeEach(() => {
   mkdirSync(TEST_DIR, { recursive: true });
   currentConfig = {
-    sandbox: { enabled: false, backend: "native" },
     services: {
       inference: {
         mode: "your-own",
@@ -142,9 +140,9 @@ beforeEach(() => {
       "image-generation": {
         mode: "your-own",
         provider: "gemini",
-        model: "gemini-2.5-flash-image",
+        model: "gemini-3.1-flash-image-preview",
       },
-      "web-search": { mode: "your-own", provider: "anthropic-native" },
+      "web-search": { mode: "your-own", provider: "inference-provider-native" },
     },
   };
 });
@@ -186,7 +184,7 @@ function createSkillOnDisk(
 // ---------------------------------------------------------------------------
 
 describe("buildSystemPrompt assistant feature flag filtering", () => {
-  test("flag OFF skill does not appear in <available_skills> section", () => {
+  test("flag OFF skill does not appear in skills catalog", () => {
     createSkillOnDisk(
       DECLARED_SKILL_ID,
       "Contacts",
@@ -201,7 +199,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       assistantFeatureFlagValues: {
         [DECLARED_FLAG_KEY]: false,
         "feature_flags.browser.enabled": true,
@@ -215,17 +212,20 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
     // browser is explicitly enabled, declared flagged skill is explicitly off
-    expect(result).toContain('id="browser"');
-    expect(result).not.toContain(`id="${DECLARED_SKILL_ID}"`);
+    expect(result).toContain("**browser**");
+    expect(result).not.toContain(`**${DECLARED_SKILL_ID}**`);
   });
 
   test("declared skills hidden when no flag overrides set (registry defaults to false)", () => {
@@ -243,7 +243,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       services: {
         inference: {
           mode: "your-own",
@@ -253,17 +252,20 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
     // Both skills declare feature flags with registry defaultEnabled: false
-    expect(result).not.toContain(`id="${DECLARED_SKILL_ID}"`);
-    expect(result).not.toContain('id="email-channel"');
+    expect(result).not.toContain(`**${DECLARED_SKILL_ID}**`);
+    expect(result).not.toContain("**email-channel**");
   });
 
   test("flagged-off skills hidden when all flags are OFF", () => {
@@ -281,7 +283,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       assistantFeatureFlagValues: {
         [DECLARED_FLAG_KEY]: false,
         "feature_flags.email-channel.enabled": false,
@@ -295,16 +296,19 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
-    expect(result).not.toContain(`id="${DECLARED_SKILL_ID}"`);
-    expect(result).not.toContain('id="email-channel"');
+    expect(result).not.toContain(`**${DECLARED_SKILL_ID}**`);
+    expect(result).not.toContain("**email-channel**");
   });
 
   test("assistantFeatureFlagValues overrides control visibility", () => {
@@ -316,7 +320,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       assistantFeatureFlagValues: { [DECLARED_FLAG_KEY]: true },
       services: {
         inference: {
@@ -327,15 +330,18 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
-    expect(result).toContain(`id="${DECLARED_SKILL_ID}"`);
+    expect(result).toContain(`**${DECLARED_SKILL_ID}**`);
   });
 
   test("persisted overrides for undeclared flags are respected", () => {
@@ -347,7 +353,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       assistantFeatureFlagValues: { "feature_flags.browser.enabled": false },
       services: {
         inference: {
@@ -358,9 +363,12 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
@@ -368,7 +376,7 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
 
     // browser declares featureFlag: "browser" and the user
     // explicitly disabled it — that override must be honored.
-    expect(result).not.toContain('id="browser"');
+    expect(result).not.toContain("**browser**");
   });
 
   test("declared flags with no persisted override use registry default", () => {
@@ -380,7 +388,6 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
     );
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       services: {
         inference: {
           mode: "your-own",
@@ -390,23 +397,25 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
     // browser is declared in the registry with defaultEnabled: true
-    expect(result).toContain('id="browser"');
+    expect(result).toContain("**browser**");
   });
 
   test("skill without featureFlag is never flag-gated", () => {
     createSkillOnDisk("my-skill", "My Skill", "A skill without feature flag");
 
     currentConfig = {
-      sandbox: { enabled: false, backend: "native" },
       services: {
         inference: {
           mode: "your-own",
@@ -416,16 +425,19 @@ describe("buildSystemPrompt assistant feature flag filtering", () => {
         "image-generation": {
           mode: "your-own",
           provider: "gemini",
-          model: "gemini-2.5-flash-image",
+          model: "gemini-3.1-flash-image-preview",
+        },
+        "web-search": {
+          mode: "your-own",
+          provider: "inference-provider-native",
         },
-        "web-search": { mode: "your-own", provider: "anthropic-native" },
       },
     };
 
     const result = buildSystemPrompt();
 
     // Skills without featureFlag declared are never gated — always pass through
-    expect(result).toContain('id="my-skill"');
+    expect(result).toContain("**my-skill**");
   });
 });
 

package/src/__tests__/attachments-store.test.ts

@@ -31,7 +31,7 @@ mock.module("../config/loader.js", () => ({
     model: "test",
     provider: "test",
     memory: { enabled: false },
-    rateLimit: { maxRequestsPerMinute: 0, maxTokensPerSession: 0 },
+    rateLimit: { maxRequestsPerMinute: 0 },
   }),
 }));
 
@@ -167,7 +167,7 @@ describe("uploadAttachment", () => {
   });
 
   test("accepts payload exactly at MAX_UPLOAD_BYTES", () => {
-    // MAX_UPLOAD_BYTES (50 MB) is divisible by 3, so (MAX/3)*4 base64 chars
+    // MAX_UPLOAD_BYTES (100 MB) is divisible by 3, so (MAX/3)*4 base64 chars
     // decodes to exactly MAX bytes with no padding.
     const exactLength = (MAX_UPLOAD_BYTES / 3) * 4;
     const exactData = "A".repeat(exactLength);

package/src/__tests__/avatar-e2e.test.ts

@@ -27,7 +27,7 @@ const geminiGenerateContentFn = mock(async () => geminiGenerateContentResult);
 
 mock.module("../config/loader.js", () => ({
   getConfig: () => ({
-    imageGenModel: "gemini-2.5-flash-image",
+    imageGenModel: "gemini-3.1-flash-image-preview",
     services: {
       inference: {
         mode: "your-own",
@@ -37,9 +37,9 @@ mock.module("../config/loader.js", () => ({
       "image-generation": {
         mode: "your-own",
         provider: "gemini",
-        model: "gemini-2.5-flash-image",
+        model: "gemini-3.1-flash-image-preview",
       },
-      "web-search": { mode: "your-own", provider: "anthropic-native" },
+      "web-search": { mode: "your-own", provider: "inference-provider-native" },
     },
   }),
 }));
@@ -47,6 +47,8 @@ mock.module("../config/loader.js", () => ({
 mock.module("../security/secure-keys.js", () => ({
   getSecureKeyAsync: async (name: string) =>
     name === "gemini" ? mockGeminiKey : null,
+  getProviderKeyAsync: async (provider: string) =>
+    provider === "gemini" ? mockGeminiKey : undefined,
 }));
 
 mock.module("../util/platform.js", () => ({

package/src/__tests__/browser-skill-endstate.test.ts

@@ -8,7 +8,6 @@ import { afterAll, beforeAll, describe, expect, mock, test } from "bun:test";
 
 mock.module("../config/loader.js", () => ({
   getConfig: () => ({
-    sandbox: { enabled: false, backend: "native" },
     assistantFeatureFlagValues: {
       "feature_flags.browser.enabled": true,
     },

package/src/__tests__/call-routes-http.test.ts

@@ -54,7 +54,7 @@ mock.module("../config/loader.js", () => ({
     model: "test",
     provider: "test",
     memory: { enabled: false },
-    rateLimit: { maxRequestsPerMinute: 0, maxTokensPerSession: 0 },
+    rateLimit: { maxRequestsPerMinute: 0 },
     secretDetection: { enabled: false },
     calls: mockCallsConfig,
   }),
@@ -62,7 +62,7 @@ mock.module("../config/loader.js", () => ({
     model: "test",
     provider: "test",
     memory: { enabled: false },
-    rateLimit: { maxRequestsPerMinute: 0, maxTokensPerSession: 0 },
+    rateLimit: { maxRequestsPerMinute: 0 },
     secretDetection: { enabled: false },
     calls: mockCallsConfig,
     ingress: {

package/src/__tests__/callback-handoff-copy.test.ts

@@ -17,7 +17,7 @@ function buildSignal(
     signalId: "test-signal-1",
     createdAt: Date.now(),
     sourceChannel: "phone",
-    sourceSessionId: "test-session-1",
+    sourceContextId: "test-session-1",
     sourceEventName: "ingress.access_request.callback_handoff",
     contextPayload: {
       requestId: "req-123",

package/src/__tests__/cancel-resolves-conversation-key.test.ts

@@ -0,0 +1,158 @@
+/**
+ * Regression test: POST /v1/conversations/:id/cancel must resolve the
+ * conversation key to the internal conversation ID before calling
+ * cancelGeneration(). Without resolveConversationId(), the cancel
+ * endpoint receives the client's local conversation key (which differs
+ * from the daemon's internal ID), fails to find the conversation, and
+ * silently ignores the cancel — leaving the stream running.
+ */
+import { mkdtempSync, realpathSync, rmSync } from "node:fs";
+import { tmpdir } from "node:os";
+import { join } from "node:path";
+import { afterAll, describe, expect, mock, test } from "bun:test";
+
+mock.module("../config/env.js", () => ({ isHttpAuthDisabled: () => true }));
+
+const testDir = realpathSync(
+  mkdtempSync(join(tmpdir(), "cancel-resolve-key-test-")),
+);
+
+mock.module("../util/platform.js", () => ({
+  getRootDir: () => testDir,
+  getDataDir: () => testDir,
+  isMacOS: () => process.platform === "darwin",
+  isLinux: () => process.platform === "linux",
+  isWindows: () => process.platform === "win32",
+  getPidPath: () => join(testDir, "test.pid"),
+  getDbPath: () => join(testDir, "test.db"),
+  getLogPath: () => join(testDir, "test.log"),
+  ensureDataDir: () => {},
+}));
+
+mock.module("../util/logger.js", () => ({
+  getLogger: () =>
+    new Proxy({} as Record<string, unknown>, {
+      get: () => () => {},
+    }),
+}));
+
+mock.module("../config/loader.js", () => ({
+  getConfig: () => ({
+    ui: {},
+    model: "test",
+    provider: "test",
+    memory: { enabled: false },
+    rateLimit: { maxRequestsPerMinute: 0 },
+    secretDetection: { enabled: false },
+    contextWindow: { maxInputTokens: 200000 },
+    services: {
+      inference: {
+        mode: "your-own",
+        provider: "anthropic",
+        model: "claude-opus-4-6",
+      },
+      "image-generation": {
+        mode: "your-own",
+        provider: "gemini",
+        model: "gemini-3.1-flash-image-preview",
+      },
+      "web-search": { mode: "your-own", provider: "inference-provider-native" },
+    },
+  }),
+}));
+
+import { getOrCreateConversation } from "../memory/conversation-key-store.js";
+import { initializeDb, resetDb } from "../memory/db.js";
+import { conversationManagementRouteDefinitions } from "../runtime/routes/conversation-management-routes.js";
+
+initializeDb();
+
+afterAll(() => {
+  resetDb();
+  try {
+    rmSync(testDir, { recursive: true, force: true });
+  } catch {
+    /* best effort */
+  }
+});
+
+describe("POST /v1/conversations/:id/cancel", () => {
+  test("resolves conversation key to internal ID before cancelling", () => {
+    // Create a conversation via key — this assigns an internal ID that
+    // differs from the key.
+    const conversationKey = "client-local-uuid-abc123";
+    const mapping = getOrCreateConversation(conversationKey);
+    const internalId = mapping.conversationId;
+
+    // Sanity: key and internal ID should differ.
+    expect(internalId).not.toBe(conversationKey);
+
+    // Track which ID cancelGeneration receives.
+    let cancelledId: string | undefined;
+    const routes = conversationManagementRouteDefinitions({
+      switchConversation: async () => null,
+      renameConversation: () => false,
+      clearAllConversations: () => 0,
+      cancelGeneration: (id) => {
+        cancelledId = id;
+        return true;
+      },
+      destroyConversation: () => {},
+      undoLastMessage: async () => null,
+      regenerateResponse: async () => null,
+    });
+
+    const cancelRoute = routes.find(
+      (r) => r.endpoint === "conversations/:id/cancel",
+    )!;
+
+    // Simulate the HTTP handler with the conversation KEY (what the
+    // macOS client sends — it uses the key, not the internal ID).
+    cancelRoute.handler({
+      params: { id: conversationKey },
+      req: new Request("http://localhost/v1/conversations/x/cancel", {
+        method: "POST",
+      }),
+      url: new URL("http://localhost/v1/conversations/x/cancel"),
+      server: undefined as never,
+      authContext: undefined as never,
+    });
+
+    // cancelGeneration must receive the INTERNAL ID, not the raw key.
+    expect(cancelledId).toBe(internalId);
+  });
+
+  test("falls back to raw ID when key is not in the mapping", () => {
+    let cancelledId: string | undefined;
+    const routes = conversationManagementRouteDefinitions({
+      switchConversation: async () => null,
+      renameConversation: () => false,
+      clearAllConversations: () => 0,
+      cancelGeneration: (id) => {
+        cancelledId = id;
+        return true;
+      },
+      destroyConversation: () => {},
+      undoLastMessage: async () => null,
+      regenerateResponse: async () => null,
+    });
+
+    const cancelRoute = routes.find(
+      (r) => r.endpoint === "conversations/:id/cancel",
+    )!;
+
+    // Use an ID that isn't a known key — should pass through as-is.
+    const directId = "direct-conversation-id";
+    cancelRoute.handler({
+      params: { id: directId },
+      req: new Request("http://localhost/v1/conversations/x/cancel", {
+        method: "POST",
+      }),
+      url: new URL("http://localhost/v1/conversations/x/cancel"),
+      server: undefined as never,
+      authContext: undefined as never,
+    });
+
+    expect(cancelledId).toBe(directId);
+  });
+});

package/src/__tests__/channel-readiness-routes.test.ts

@@ -19,7 +19,6 @@ let mockPrimaryInboxAddress: string | undefined;
 mock.module("../calls/twilio-rest.js", () => ({
   hasTwilioCredentials: () => mockHasTwilioCredentials,
   getPhoneNumberSid: async () => null,
-  getTollFreeVerificationStatus: async () => null,
 }));
 
 mock.module("../config/env.js", () => ({}));

package/src/__tests__/channel-readiness-service.test.ts

@@ -13,7 +13,6 @@ let mockGatewayHealth = {
 
 mock.module("../calls/twilio-rest.js", () => ({
   getPhoneNumberSid: async () => null,
-  getTollFreeVerificationStatus: async () => null,
   getTwilioCredentials: () => ({
     accountSid: "AC_test",
     authToken: "test-auth-token",