arul28 · arul28 · Sep 4, 2026 · Sep 4, 2026 · Sep 4, 2026 · Sep 4, 2026
@@ -1152,8 +1152,8 @@ export async function createAdeRuntime(args: {
     const gitService = createGitOperationsService({
       laneService,
       operationService,
-      projectConfigService,
       aiIntegrationService,
+      sessionService,
       logger
     });
 

@@ -4081,6 +4081,7 @@ describe("ADE CLI", () => {
     if (help.kind === "help") {
       expect(help.text).toContain("ade chat note");
       expect(help.text).toContain("ade chat ask");
+      expect(help.text).toContain("ade chat generate-names");
       expect(help.text).toContain("ade chat demote");
       expect(help.text).toContain("ade chat promote");
       // Settling is user-/PR-merge-driven only; the help must say so rather
@@ -4117,6 +4118,40 @@ describe("ADE CLI", () => {
     },
   );
 
+  it("passes --session through for chat generate-names and defaults to all fields", () => {
+    const plan = expectExecutePlan(buildCliPlan([
+      "chat",
+      "generate-names",
+      "--session",
+      "session-x",
+    ]));
+    expect(plan.steps[0]?.params).toMatchObject({
+      arguments: {
+        domain: "chat",
+        action: "regenerateSessionMetadata",
+        args: { sessionId: "session-x" },
+      },
+    });
+  });
+
+  it("forwards generate-names field flags", () => {
+    const plan = expectExecutePlan(buildCliPlan([
+      "chat",
+      "generate-names",
+      "--title",
+      "--status",
+      "--session",
+      "session-x",
+    ]));
+    expect(plan.steps[0]?.params).toMatchObject({
+      arguments: {
+        domain: "chat",
+        action: "regenerateSessionMetadata",
+        args: { sessionId: "session-x", fields: ["title", "statusLine"] },
+      },
+    });
+  });
+
   describe("session lifecycle commands", () => {
     const NOW = Date.parse("2026-07-26T12:00:00.000Z");
 

@@ -2124,6 +2124,9 @@ const HELP_BY_COMMAND: Record<string, string> = {
                                                     'note' and 'ask' default to the caller and accept --session <id>.
                                                     'chat settle' / 'chat unsettle' were removed: only the user (or a
                                                     merged PR) settles a session — report your outcome with 'chat note'.
+    $ ade chat generate-names                       Regenerate chat title, lane name, and status line
+    $ ade chat generate-names --title --status      Limit to those fields; omit flags for all three
+                                                    Also: ade chat update --title, ade lanes rename.
     $ ade chat steer <session> --personal --text "focus on the tradeoffs"
     $ ade chat models --personal --provider codex
     $ ade chat update <session> --personal --title "Trip planning"
@@ -7575,7 +7578,8 @@ function buildChatPlan(args: string[]): CliPlan {
     : null;
   // `ask` / `note` take free text, not a session positional — they default to
   // the caller's own $ADE_CHAT_SESSION_ID and accept --session <id>.
-  const selfLifecycleSub = sub === "ask" || sub === "note";
+  const selfLifecycleSub = sub === "ask" || sub === "note"
+    || sub === "generate-names" || sub === "generate_names" || sub === "names";
   const explicitSessionId = readValue(args, ["--session", "--session-id"]);
   const sessionId =
     explicitSessionId ??
@@ -7621,6 +7625,24 @@ function buildChatPlan(args: string[]): CliPlan {
       ],
     };
   }
+  if (sub === "generate-names" || sub === "generate_names" || sub === "names") {
+    const fields: string[] = [];
+    if (readFlag(args, ["--title"])) fields.push("title");
+    if (readFlag(args, ["--lane", "--lane-name"])) fields.push("laneName");
+    if (readFlag(args, ["--status", "--status-line"])) fields.push("statusLine");
+    return {
+      kind: "execute",
+      label: "chat generate-names",
+      steps: [
+        actionStep(
+          "result",
+          "chat",
+          "regenerateSessionMetadata",
+          withSession(fields.length ? { fields } : {}),
+        ),
+      ],
+    };
+  }
   if (sub === "list" || sub === "ls") {
     const includeArchived = readFlag(args, ["--archived", "--include-archived"]);
     const excludeArchived = readFlag(args, [

@@ -222,6 +222,17 @@ sessions. Running a removed command fails with that explanation.
 What to do instead when you finish: say so in your final message, and use
 `ade chat note "<one-line status>"` to leave a durable status line on the Work
 row. If you are blocked, `ade chat ask "<question>"` raises the row's hand.
+Update the note along the way as the state changes; do not wait until the end.
+
+If you realize the lane, branch, or chat name is wrong, rename it rather than
+living with a bad label:
+
+```bash
+ade chat generate-names                      # title, lane name, and status line
+ade chat generate-names --title --status     # subset of fields
+ade chat update --title "Better chat title"  # defaults to $ADE_CHAT_SESSION_ID
+ade lanes rename <lane> --name "Better name"
+```
 
 #### What `note` and `ask` do to the Work row
 

@@ -245,7 +245,7 @@
 import { localIpcListenOptions } from "../../../ade-cli/src/services/runtime/localIpcListenOptions";
 import { normalizeProjectRootPath } from "../../../ade-cli/src/services/projects/projectRoots";
 import {
  ACCOUNT_SESSION_CREDENTIAL_KEY,
  getSignedInAccountAccessToken,
  type AccountAuthService,
 } from "../../../ade-cli/src/services/account/accountAuthService";
@@ -4007,8 +4007,8 @@
     const gitService = createGitOperationsService({
       laneService,
       operationService,
-      projectConfigService,
       aiIntegrationService,
+      sessionService,
       logger,
       onWorktreeChanged: ({ laneId, reason }) => {
         jobEngine.onLaneDirtyChanged({ laneId, reason });

@@ -263,6 +263,7 @@ const SESSION_INTELLIGENCE_TASK_TYPES: ReadonlySet<AiTaskType> = new Set([
 const EXPLICIT_MODEL_ONLY_TASK_TYPES: ReadonlySet<AiTaskType> = new Set([
   ...SESSION_INTELLIGENCE_TASK_TYPES,
   "context_compaction",
+  "commit_message",
 ]);
 
 export function readConfiguredFeatureModel(aiConfig: unknown, feature: AiFeatureKey): string | null {
@@ -1706,7 +1707,8 @@ export function createAiIntegrationService(args: {
       throw new Error("No AI provider is available. Install and authenticate Claude Code and/or Codex CLI.");
     }
 
-    if (args.taskType !== "session_title" && !getFeatureFlag(args.feature)) {
+    // Titles and commit messages always run. Settings no longer gates them.
+    if (args.taskType !== "session_title" && args.taskType !== "commit_message" && !getFeatureFlag(args.feature)) {
       logger.warn("ai.task.skipped_feature_disabled", {
         requestId,
         taskType: args.taskType,

@@ -1636,20 +1636,14 @@ export function createCtoOperatorTools(deps: CtoOperatorToolDeps): Record<string
   });
 
   tools.listConflictProposals = tool({
-    description: "List AI-generated conflict resolution proposals for a lane.",
+    description: "List stored conflict resolution proposals for a lane.",
     inputSchema: z.object({ laneId: z.string().min(1) }),
     execute: ({ laneId }) => conflictGuard(async () => {
       const proposals = await deps.conflictService!.listProposals({ laneId });
       return { count: proposals.length, proposals };
     }),
   });
 
-  tools.requestConflictProposal = tool({
-    description: "Request an AI-generated resolution for a specific conflict.",
-    inputSchema: z.object({ laneId: z.string().min(1), filePath: z.string().optional() }),
-    execute: ({ laneId, filePath }) => conflictGuard(() => deps.conflictService!.requestProposal({ laneId, filePath: filePath?.trim() || undefined })),
-  });
-
   tools.applyConflictProposal = tool({
     description: "Apply an AI-generated conflict resolution proposal.",
     inputSchema: z.object({ laneId: z.string().min(1), proposalId: z.string().min(1) }),

@@ -1541,6 +1541,7 @@ function createMockSessionService() {
       if (row) row.statusNote = note;
       return Boolean(row);
     }),
+    getStatusNoteUpdatedAt: vi.fn(() => null),
     setHeadShaStart: vi.fn(),
     setHeadShaEnd: vi.fn(),
     setLastOutputPreview: vi.fn(),
@@ -1784,6 +1785,7 @@ function createService(overrides: Record<string, unknown> = {}) {
     aiIntegrationService: aiIntegrationService as any,
     logger: logger as any,
     appVersion: "0.0.1-test",
+    nativeTitleWaitMs: 0,
     getDirtyFileTextForPath: () => undefined,
     ...overrides,
   });
@@ -13636,15 +13638,15 @@ describe("createAgentChatService", () => {
           && event.event.status === "spawn_completion_delivery_failed"
         )).toBe(true);
       }, { timeout: 2_500 });
-      expect(logger.warn).toHaveBeenCalledTimes(3);
-      expect(logger.warn).toHaveBeenLastCalledWith(
-        "agent_chat.spawn_completion_delivery_failed",
-        expect.objectContaining({
-          childSessionId: child.id,
-          parentSessionId: parent.id,
-          attempt: 3,
-        }),
+      const deliveryWarnings = logger.warn.mock.calls.filter(
+        ([message]) => message === "agent_chat.spawn_completion_delivery_failed",
       );
+      expect(deliveryWarnings).toHaveLength(3);
+      expect(deliveryWarnings[2]?.[1]).toEqual(expect.objectContaining({
+        childSessionId: child.id,
+        parentSessionId: parent.id,
+        attempt: 3,
+      }));
     });
 
     it("rejects the legacy silent spawn type for new child chats", async () => {
@@ -16022,6 +16024,8 @@ describe("createAgentChatService", () => {
 
       const { service, sessionService } = createService({
         onEvent: (event: AgentChatEventEnvelope) => events.push(event),
+        // These tests prove native Claude titles win during the wait window.
+        nativeTitleWaitMs: 250,
         projectConfigService: {
           get: vi.fn(() => ({
             effective: {
@@ -16067,9 +16071,8 @@ describe("createAgentChatService", () => {
         info: { summary: prompt, firstPrompt: prompt },
         firstPrompt: prompt,
       });
-      // Give the fire-and-forget adopt a beat, then confirm the title stayed default.
-      await new Promise((resolve) => setTimeout(resolve, 20));
-      expect(sessionService.get(session.id)?.title).toBe("Claude Chat");
+      // Echoed SDK summaries are skipped; ADE names the chat after the wait.
+      await waitForSessionTitle(sessionService, session.id, "Fix Update Modal Flow");
     });
 
     it("does not adopt when the session is manually named", async () => {
@@ -17942,10 +17945,17 @@ describe("createAgentChatService", () => {
           event.event.type === "done",
       );
 
-      // Give auto-title a chance to fire (it's a void promise)
+      // Give auto-title / idle status-line a chance to fire (void promises)
       await new Promise((resolve) => setTimeout(resolve, 50));
 
-      expect(aiIntegrationService.summarizeTerminal).not.toHaveBeenCalled();
+      expect(sessionService.get(session.id)?.title).toBe("My Title");
+      expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalled();
+      expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalledWith(
+        expect.objectContaining({
+          prompt: expect.stringContaining("Write a short statusLine"),
+          systemPrompt: expect.stringContaining("Copy these current values unchanged: chatTitle, laneName"),
+        }),
+      );
     });
 
     it("does not clobber a manual rename that lands while auto-titling is in flight", async () => {
@@ -47149,21 +47159,27 @@ describe("suggestLaneNameFromPrompt", () => {
     );
   });
 
-  it("uses the deterministic prompt fallback when title generation is disabled", async () => {
+  it("still names with AI when title generation is disabled in Settings", async () => {
     vi.mocked(detectAllAuth).mockResolvedValue([
       { type: "cli-subscription" as any, cli: "claude", authenticated: true, path: "/usr/bin/claude", verified: true },
     ]);
 
     const { service, aiIntegrationService } = createSuggestService({ titleGenerationEnabled: false });
+    vi.mocked(aiIntegrationService.summarizeTerminal).mockResolvedValue({
+      text: "Login Bug Fix",
+      inputTokens: 10,
+      outputTokens: 5,
+    } as any);
     const result = await service.suggestLaneNameFromPrompt({
       prompt: "Fix the authentication login failure in the dashboard",
       modelId: "anthropic/claude-haiku-4-5",
+      provider: "claude",
       laneId: "lane-1",
       fallbackName: "chat-20260514-010203",
     });
 
-    expect(result).toBe("fix-authentication-login-failure-dashboard");
-    expect(aiIntegrationService.summarizeTerminal).not.toHaveBeenCalled();
+    expect(result).toBe("login-bug-fix");
+    expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalled();
   });
 
   it("preserves the generated suffix when the prompt fallback is generic", async () => {
@@ -47198,7 +47214,7 @@ describe("suggestLaneNameFromPrompt", () => {
     expect(result).toBe("login-bug-fix");
   });
 
-  it("prefers the configured title model over the requested composer model", async () => {
+  it("prefers the cheap helper for the ADE provider over the requested session model", async () => {
     vi.mocked(detectAllAuth).mockResolvedValue([
       { type: "cli-subscription" as any, cli: "codex", authenticated: true, path: "/usr/bin/codex", verified: true },
     ]);
@@ -47212,13 +47228,14 @@ describe("suggestLaneNameFromPrompt", () => {
     const result = await service.suggestLaneNameFromPrompt({
       prompt: "Fix auto create lane routing and naming",
       modelId: "openai/gpt-5.5",
+      provider: "codex",
       laneId: "lane-1",
       fallbackName: "chat-20260514-010203",
     });
 
     expect(result).toBe("auto-create-lane-fix");
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenNthCalledWith(1, expect.objectContaining({
-      model: "openai/gpt-5.4-mini",
+      model: "openai/gpt-5.6-luna",
       taskType: "session_title",
     }));
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalledTimes(1);
@@ -47480,7 +47497,7 @@ describe("suggestLaneNameFromPrompt", () => {
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalledTimes(1);
   });
 
-  it("uses the launched chat model when the title setting answers unusably", async () => {
+  it("uses the launched chat model when the cheap helper answers unusably", async () => {
     vi.mocked(detectAllAuth).mockResolvedValue([
       { type: "cli-subscription" as any, cli: "codex", authenticated: true, path: "/usr/bin/codex", verified: true },
     ]);
@@ -47496,6 +47513,7 @@ describe("suggestLaneNameFromPrompt", () => {
     const result = await service.generateAutoLaneIdentity({
       prompt: "The Claude auth login button hangs after OAuth redirects.",
       modelId: "openai/gpt-5.4",
+      provider: "codex",
       laneId: "lane-1",
       temporaryBranch: "ade/1a2b3c4d",
     });
@@ -47506,15 +47524,15 @@ describe("suggestLaneNameFromPrompt", () => {
       source: "ai",
     });
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenNthCalledWith(1, expect.objectContaining({
-      model: "openai/gpt-5.4-mini",
+      model: "openai/gpt-5.6-luna",
     }));
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenNthCalledWith(2, expect.objectContaining({
       model: "openai/gpt-5.4",
     }));
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenCalledTimes(2);
   });
 
-  it("uses the configured naming model before the launched model", async () => {
+  it("uses the cheap helper before the launched model", async () => {
     vi.mocked(detectAllAuth).mockResolvedValue([
       { type: "cli-subscription" as any, cli: "codex", authenticated: true, path: "/usr/bin/codex", verified: true },
     ]);
@@ -47526,12 +47544,13 @@ describe("suggestLaneNameFromPrompt", () => {
     await service.generateAutoLaneIdentity({
       prompt: "Rename automatic lanes",
       modelId: "openai/gpt-5.4",
+      provider: "codex",
       laneId: "lane-1",
       temporaryBranch: "ade/1a2b3c4d",
     });
 
     expect(aiIntegrationService.summarizeTerminal).toHaveBeenNthCalledWith(1, expect.objectContaining({
-      model: "openai/gpt-5.4-mini",
+      model: "openai/gpt-5.6-luna",
     }));
   });