feat: automated task loop panel with per-task conversation orchestration

Implements issue #179. Adds a Task Loop panel that imports hikari-tasks.json
files and automatically runs Claude Code through each task in sequence, one
per conversation tab. Includes Start/Pause/Stop/Resume controls, real-time
status indicators per task, and a StatusBar badge. WSL UNC paths from the
file picker are normalised to Unix paths at import time.

src/lib/stores/taskLoop.test.ts

@@ -0,0 +1,127 @@
+import { describe, it, expect } from "vitest";
+import {
+  findNextPendingIndex,
+  countByStatus,
+  buildTaskPrompt,
+  normalizeToUnixPath,
+  type TaskLoopTask,
+} from "./taskLoop";
+
+const makeTask = (id: string, status: TaskLoopTask["status"] = "pending"): TaskLoopTask => ({
+  id,
+  title: `Task ${id}`,
+  prompt: `Do the thing for ${id}`,
+  priority: "medium",
+  status,
+});
+
+describe("findNextPendingIndex", () => {
+  it("returns -1 for an empty list", () => {
+    expect(findNextPendingIndex([])).toBe(-1);
+  });
+
+  it("returns 0 when the first task is pending", () => {
+    const tasks = [makeTask("1", "pending"), makeTask("2", "pending")];
+    expect(findNextPendingIndex(tasks)).toBe(0);
+  });
+
+  it("skips completed and failed tasks to find the next pending", () => {
+    const tasks = [
+      makeTask("1", "completed"),
+      makeTask("2", "failed"),
+      makeTask("3", "pending"),
+      makeTask("4", "pending"),
+    ];
+    expect(findNextPendingIndex(tasks)).toBe(2);
+  });
+
+  it("returns -1 when all tasks are completed", () => {
+    const tasks = [makeTask("1", "completed"), makeTask("2", "completed")];
+    expect(findNextPendingIndex(tasks)).toBe(-1);
+  });
+
+  it("skips running tasks", () => {
+    const tasks = [makeTask("1", "running"), makeTask("2", "pending")];
+    expect(findNextPendingIndex(tasks)).toBe(1);
+  });
+});
+
+describe("countByStatus", () => {
+  it("returns 0 for an empty list", () => {
+    expect(countByStatus([], "pending")).toBe(0);
+  });
+
+  it("counts only tasks with the specified status", () => {
+    const tasks = [
+      makeTask("1", "pending"),
+      makeTask("2", "completed"),
+      makeTask("3", "pending"),
+      makeTask("4", "failed"),
+    ];
+    expect(countByStatus(tasks, "pending")).toBe(2);
+    expect(countByStatus(tasks, "completed")).toBe(1);
+    expect(countByStatus(tasks, "failed")).toBe(1);
+    expect(countByStatus(tasks, "running")).toBe(0);
+  });
+
+  it("counts all tasks when all have the same status", () => {
+    const tasks = [makeTask("1", "completed"), makeTask("2", "completed")];
+    expect(countByStatus(tasks, "completed")).toBe(2);
+  });
+});
+
+describe("buildTaskPrompt", () => {
+  it("includes the task number and total", () => {
+    const task = makeTask("1");
+    const result = buildTaskPrompt(task, 1, 5);
+    expect(result).toContain("1/5");
+  });
+
+  it("includes the task title", () => {
+    const task = makeTask("abc");
+    const result = buildTaskPrompt(task, 2, 3);
+    expect(result).toContain("Task abc");
+  });
+
+  it("includes the task prompt", () => {
+    const task = makeTask("x");
+    const result = buildTaskPrompt(task, 1, 1);
+    expect(result).toContain("Do the thing for x");
+  });
+
+  it("labels the output as an automated task loop entry", () => {
+    const task = makeTask("1");
+    const result = buildTaskPrompt(task, 1, 1);
+    expect(result).toContain("Automated Task Loop");
+  });
+});
+
+describe("normalizeToUnixPath", () => {
+  it("converts a WSL UNC path with wsl.localhost to a Unix path", () => {
+    expect(
+      normalizeToUnixPath("\\\\wsl.localhost\\Ubuntu\\home\\naomi\\code\\temp\\file.json")
+    ).toBe("/home/naomi/code/temp/file.json");
+  });
+
+  it("converts a WSL UNC path with wsl$ (legacy) to a Unix path", () => {
+    expect(normalizeToUnixPath("\\\\wsl$\\Ubuntu\\home\\naomi\\file.json")).toBe(
+      "/home/naomi/file.json"
+    );
+  });
+
+  it("handles forward-slash UNC paths produced by some tools", () => {
+    expect(normalizeToUnixPath("//wsl.localhost/Ubuntu/home/naomi/file.json")).toBe(
+      "/home/naomi/file.json"
+    );
+  });
+
+  it("leaves a plain Unix path unchanged", () => {
+    expect(normalizeToUnixPath("/home/naomi/code/temp/file.json")).toBe(
+      "/home/naomi/code/temp/file.json"
+    );
+  });
+
+  it("leaves an empty string unchanged", () => {
+    expect(normalizeToUnixPath("")).toBe("");
+  });
+});