feat: make chat-style tasks the default experience (#11101)

server/src/services/built-in-agents.ts

@@ -471,6 +471,12 @@ const DEFINITIONS = validateBuiltInAgentDefinitions([
 
 const DEFINITIONS_BY_KEY = new Map(DEFINITIONS.map((definition) => [definition.key, definition]));
 
+// Bundled built-in agents that should be provisioned automatically when a
+// company is created (and re-ensured on startup reconcile). Empty by default so
+// a new user starts clean — the Reflection Coach and Summarizer are opt-in, not
+// seeded. Add a definition key here to restore automatic provisioning.
+const AUTO_PROVISION_ON_COMPANY_CREATE_KEYS = new Set<string>([]);
+
 const ROOT_AGENT_DEFAULT_CHANGE_GRANTS: PermissionKey[] = ["agents:configure", "skills:create"];
 const BUILT_IN_AGENT_DEFAULT_GRANTS: Record<string, PermissionKey[]> = {
   "reflection-coach": ["agents:suggest-changes", "skills:suggest-changes"],
@@ -1917,7 +1923,17 @@ export function builtInAgentService(db: Db) {
     const company = await ensureCompany(companyId);
     let autoEnsured = 0;
     let pendingApprovals = 0;
+    // A fresh company starts with only its own lead agent — the Reflection
+    // Coach and Summarizer are no longer auto-created for new users. They stay
+    // available to enable on demand (via ensure / provision / the built-in
+    // bundle panel). We still reconcile any bundled agent that already exists
+    // (e.g. one an operator enabled) so its instructions/skill/routine keep
+    // tracking stock. Add a key to AUTO_PROVISION_ON_COMPANY_CREATE_KEYS to
+    // restore automatic creation for that definition.
     for (const definition of DEFINITIONS.filter((entry) => entry.bundle)) {
+      const existing = await findSingleAgent(companyId, definition);
+      const shouldProvision = existing !== null || AUTO_PROVISION_ON_COMPANY_CREATE_KEYS.has(definition.key);
+      if (!shouldProvision) continue;
       if (company.requireBoardApprovalForNewAgents) {
         const result = await provision(companyId, definition.key);
         if (result.approval) pendingApprovals += 1;

server/src/services/instance-settings.ts

@@ -214,7 +214,7 @@ export function normalizeExperimentalSettings(raw: unknown): InstanceExperimenta
       enablePipelines: parsed.data.enablePipelines ?? false,
       enableCases: parsed.data.enableCases ?? false,
       enableConferenceRoomChat: parsed.data.enableConferenceRoomChat ?? false,
-      enableTaskChatRedesign: parsed.data.enableTaskChatRedesign ?? false,
+      enableClassicTaskInterface: parsed.data.enableClassicTaskInterface ?? false,
       enableIssuePlanDecompositions: parsed.data.enableIssuePlanDecompositions ?? false,
       enableExperimentalFileViewer: parsed.data.enableExperimentalFileViewer ?? false,
       enableTaskWatchdogs: parsed.data.enableTaskWatchdogs ?? false,
@@ -250,7 +250,7 @@ export function normalizeExperimentalSettings(raw: unknown): InstanceExperimenta
     enablePipelines: false,
     enableCases: false,
     enableConferenceRoomChat: false,
-    enableTaskChatRedesign: false,
+    enableClassicTaskInterface: false,
     enableTaskWatchdogs: false,
     enableIssuePlanDecompositions: false,
     enableExperimentalFileViewer: false,

server/src/services/issue-thread-interactions.ts

@@ -565,6 +565,21 @@ function buildSupersededByNewerRequestResult(replacementInteractionId: string) {
   } as const;
 }
 
+// An agent that posts a fresh ask_user_questions while its own earlier ones on
+// the same issue are still pending has replaced them — the newer card carries
+// the real ask, so the stale siblings auto-expire (PAP-437). Mirrors the
+// `superseded_by_comment` shape (ask_user_questions results key expiry off
+// `expirationReason`, not `outcome`) so the UI can hide them cleanly.
+function buildSupersededByNewerInteractionResult(replacementInteractionId: string) {
+  return {
+    version: 1,
+    answers: [],
+    expirationReason: "superseded_by_newer_interaction",
+    supersededByInteractionId: replacementInteractionId,
+    summaryMarkdown: null,
+  } as const;
+}
+
 function buildAdministrativeOutcomeResult(
   row: IssueThreadInteractionRow,
   outcome: "withdrawn" | "issue_closed" | "addressee_deleted",
@@ -1888,16 +1903,27 @@ export function issueThreadInteractionService(db: Db, opts: IssueThreadInteracti
             })
             .returning();
 
-          if (data.kind !== "request_confirmation" || !actor.agentId) {
+          // An agent replacing its own still-pending card supersedes the older
+          // one so the thread never accumulates stale sibling cards. This covers
+          // request_confirmation drafts and ask_user_questions (PAP-437: probe
+          // question cards that agents never withdrew). Each kind keeps its own
+          // result shape. Scoped strictly to the same agent + issue + kind, so
+          // other agents' or other kinds' pending cards are untouched.
+          const canSupersedeSiblingCards =
+            data.kind === "request_confirmation" || data.kind === "ask_user_questions";
+          if (!actor.agentId || !canSupersedeSiblingCards) {
             return { row, supersededRows: [] };
           }
 
           const now = new Date();
+          const supersededResult = data.kind === "ask_user_questions"
+            ? buildSupersededByNewerInteractionResult(row.id)
+            : buildSupersededByNewerRequestResult(row.id);
           const supersededRows = await tx
             .update(issueThreadInteractions)
             .set({
               status: "expired",
-              result: buildSupersededByNewerRequestResult(row.id),
+              result: supersededResult,
               resolvedByAgentId: actor.agentId,
               resolvedByUserId: actor.userId ?? null,
               resolvedAt: now,

server/src/services/onboarding-greeting.test.ts

@@ -0,0 +1,43 @@
+import { describe, expect, it } from "vitest";
+import { buildOnboardingGreeting } from "./onboarding-greeting.js";
+
+describe("buildOnboardingGreeting", () => {
+  it("introduces the agent by name as the user's first teammate and reflects the goals", () => {
+    const greeting = buildOnboardingGreeting({
+      agentName: "Nova",
+      teamName: "Acme",
+      goals: "Launch a marketplace for local makers.",
+    });
+
+    expect(greeting).toContain(
+      "Welcome! I'm Nova, your first agent teammate on Paperclip.",
+    );
+    expect(greeting).toContain("Here's what I understand you're aiming for:");
+    expect(greeting).toContain("> Launch a marketplace for local makers.");
+    expect(greeting).toContain("propose a team of agents");
+    expect(greeting).toContain("few focused questions");
+  });
+
+  it("falls back to a generic teammate intro when no agent name is set", () => {
+    const greeting = buildOnboardingGreeting({ agentName: null, goals: null });
+
+    expect(greeting).toContain(
+      "Welcome! I'm your first agent teammate on Paperclip.",
+    );
+  });
+
+  it("collapses whitespace in the reflected goals", () => {
+    const greeting = buildOnboardingGreeting({
+      goals: "  Build\n\n  a  SaaS product.  ",
+    });
+
+    expect(greeting).toContain("> Build a SaaS product.");
+  });
+
+  it("omits the reflect-back block when no goals are provided", () => {
+    const greeting = buildOnboardingGreeting({ agentName: "Nova", goals: null });
+
+    expect(greeting).not.toContain("aiming for");
+    expect(greeting).toContain("propose a team of agents");
+  });
+});

server/src/services/onboarding-greeting.ts

@@ -0,0 +1,39 @@
+// Deterministic, template-driven greeting seeded as an agent-authored comment on
+// the onboarding first task. No LLM call: it reflects back the onboarding context
+// (team name + goals) so the user lands on a waiting greeting instead of a
+// right-aligned "user" bubble showing the agent's own seeded instructions.
+
+export const ONBOARDING_GREETING_AUTHORIZATION_REASON = "onboarding first-task greeting";
+
+export function buildOnboardingGreeting(input: {
+  agentName?: string | null;
+  teamName?: string | null;
+  goals?: string | null;
+}): string {
+  const agentName = input.agentName?.trim();
+  const goals = input.goals?.replace(/\s+/g, " ").trim();
+
+  // Introduce the agent by the name the user chose in onboarding when we have
+  // it, so the first message reads as coming from *their* first teammate rather
+  // than a generic agent. Fall back to the generic phrasing otherwise.
+  const identity = agentName
+    ? `Welcome! I'm ${agentName}, your first agent teammate on Paperclip.`
+    : "Welcome! I'm your first agent teammate on Paperclip.";
+
+  const lines: string[] = [];
+  lines.push(identity);
+
+  if (goals) {
+    lines.push("");
+    lines.push("Here's what I understand you're aiming for:");
+    lines.push("");
+    lines.push(`> ${goals}`);
+  }
+
+  lines.push("");
+  lines.push(
+    "I want to gather more context so I can come up with a plan and propose a team of agents to help execute it. I'm putting together a few focused questions so we can settle on a concrete goal to tackle first. Please give me one moment...",
+  );
+
+  return lines.join("\n");
+}