feat(sidebar): multi-select tasks with bulk archive

[fercgomes]

Summary

• Cmd/Ctrl-click toggles a task in the selection; shift-click extends a range from the last clicked anchor.
• Right-clicking a task while 2+ are selected opens a bulk context menu with "Archive N tasks" (confirm dialog). Right-clicking outside the selection clears it and falls through to the single-task menu.
• Escape clears the selection (ignored when focus is in an input/textarea/contenteditable).
• Selection lives in a small Zustand store (taskSelectionStore) modeled after the inbox pattern, with a prune effect that drops ids no longer visible when filters change.
• archiveTaskImperative gets a sibling archiveTasksImperative helper used by both the new bulk path and the existing "Archive prior tasks" action.

Test plan

• Plain click navigates and clears any selection
• Cmd/Ctrl-click toggles selection without navigating
• Shift-click selects a contiguous range across pinned + chronological tasks
• Right-click on a selected task with 2+ selected opens the bulk menu and "Archive N tasks" archives all of them
• Right-click on an unselected task with a selection clears the selection and shows the normal single-task menu
• Escape clears the selection; typing in an input does not
• Hover toolbar (pin/archive icons) is hidden while a multi-selection is active
• Changing filters (e.g. organize/sort/visibility) prunes any selected ids that are no longer visible
• pnpm --filter code typecheck, biome, and vitest (43 tests in sidebar/tasks) all pass

[greptile-apps]

Prompt To Fix All With AI

Fix the following 3 code review issues. Work through them one at a time, proposing concise fixes.

---

### Issue 1 of 3
apps/code/src/renderer/features/sidebar/stores/taskSelectionStore.test.ts:90-112
The "forward range" and "backward range" tests are symmetric cases testing the same logical property (range is order-independent) and are a natural fit for a parameterised test. Per the team's simplicity rules, we prefer `it.each` for cases like this.

```suggestion
    it.each([
      { direction: "forward", anchor: "t2", target: "t4" },
      { direction: "backward", anchor: "t4", target: "t2" },
    ])(
      "selects a $direction range from anchor to target",
      ({ anchor, target }) => {
        useTaskSelectionStore.setState({ lastClickedId: anchor });

        useTaskSelectionStore.getState().selectRange(target, orderedIds);

        expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual([
          "t2",
          "t3",
          "t4",
        ]);
      },
    );
```

### Issue 2 of 3
apps/code/src/renderer/features/sidebar/stores/taskSelectionStore.test.ts:130-142
These two cases both assert "selects just the target" with the only difference being whether `lastClickedId` is absent vs. points at an id not in the ordered list. They're a clean `it.each` opportunity.

```suggestion
    it.each([
      { case: "no anchor", lastClickedId: null },
      { case: "anchor not in ordered list", lastClickedId: "t99" },
    ])(
      "selects just the target when $case",
      ({ lastClickedId }) => {
        if (lastClickedId) {
          useTaskSelectionStore.setState({ lastClickedId });
        }

        useTaskSelectionStore.getState().selectRange("t3", orderedIds);

        expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual(["t3"]);
      },
    );
```

### Issue 3 of 3
apps/code/src/main/services/context-menu/schemas.ts:13-15
The schema uses `positive()` (≥ 1), but the context menu label renders `"Archive ${taskCount} tasks"` — if `taskCount` is 1 that reads "Archive 1 tasks". Changing to `.min(2)` also makes the schema self-document that this endpoint is intentionally bulk-only.

```suggestion
export const bulkTaskContextMenuInput = z.object({
  taskCount: z.number().int().min(2),
});
```

_{Reviews (1): Last reviewed commit: "fix(sidebar): shift-click range follows ..." | Re-trigger Greptile}

[greptile-apps]

The "forward range" and "backward range" tests are symmetric cases testing the same logical property (range is order-independent) and are a natural fit for a parameterised test. Per the team's simplicity rules, we prefer it.each for cases like this.

Suggested change

	it("selects a forward range from anchor to target", () => {
	useTaskSelectionStore.setState({ lastClickedId: "t2" });
	useTaskSelectionStore.getState().selectRange("t4", orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual([
	"t2",
	"t3",
	"t4",
	]);
	});
	it("selects a backward range from anchor to target", () => {
	useTaskSelectionStore.setState({ lastClickedId: "t4" });
	useTaskSelectionStore.getState().selectRange("t2", orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual([
	"t2",
	"t3",
	"t4",
	]);
	});
	it.each([
	{ direction: "forward", anchor: "t2", target: "t4" },
	{ direction: "backward", anchor: "t4", target: "t2" },
	])(
	"selects a $direction range from anchor to target",
	({ anchor, target }) => {
	useTaskSelectionStore.setState({ lastClickedId: anchor });
	useTaskSelectionStore.getState().selectRange(target, orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual([
	"t2",
	"t3",
	"t4",
	]);
	},
	);

Prompt To Fix With AI

This is a comment left during a code review.
Path: apps/code/src/renderer/features/sidebar/stores/taskSelectionStore.test.ts
Line: 90-112

Comment:
The "forward range" and "backward range" tests are symmetric cases testing the same logical property (range is order-independent) and are a natural fit for a parameterised test. Per the team's simplicity rules, we prefer `it.each` for cases like this.

```suggestion
    it.each([
      { direction: "forward", anchor: "t2", target: "t4" },
      { direction: "backward", anchor: "t4", target: "t2" },
    ])(
      "selects a $direction range from anchor to target",
      ({ anchor, target }) => {
        useTaskSelectionStore.setState({ lastClickedId: anchor });

        useTaskSelectionStore.getState().selectRange(target, orderedIds);

        expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual([
          "t2",
          "t3",
          "t4",
        ]);
      },
    );
```

How can I resolve this? If you propose a fix, please make it concise.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[greptile-apps]

These two cases both assert "selects just the target" with the only difference being whether lastClickedId is absent vs. points at an id not in the ordered list. They're a clean it.each opportunity.

Suggested change

	it("selects just the target when there is no anchor", () => {
	useTaskSelectionStore.getState().selectRange("t3", orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual(["t3"]);
	});
	it("selects just the target when anchor is not in the ordered list", () => {
	useTaskSelectionStore.setState({ lastClickedId: "t99" });
	useTaskSelectionStore.getState().selectRange("t3", orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual(["t3"]);
	});
	it.each([
	{ case: "no anchor", lastClickedId: null },
	{ case: "anchor not in ordered list", lastClickedId: "t99" },
	])(
	"selects just the target when $case",
	({ lastClickedId }) => {
	if (lastClickedId) {
	useTaskSelectionStore.setState({ lastClickedId });
	}
	useTaskSelectionStore.getState().selectRange("t3", orderedIds);
	expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual(["t3"]);
	},
	);

Prompt To Fix With AI

This is a comment left during a code review.
Path: apps/code/src/renderer/features/sidebar/stores/taskSelectionStore.test.ts
Line: 130-142

Comment:
These two cases both assert "selects just the target" with the only difference being whether `lastClickedId` is absent vs. points at an id not in the ordered list. They're a clean `it.each` opportunity.

```suggestion
    it.each([
      { case: "no anchor", lastClickedId: null },
      { case: "anchor not in ordered list", lastClickedId: "t99" },
    ])(
      "selects just the target when $case",
      ({ lastClickedId }) => {
        if (lastClickedId) {
          useTaskSelectionStore.setState({ lastClickedId });
        }

        useTaskSelectionStore.getState().selectRange("t3", orderedIds);

        expect(useTaskSelectionStore.getState().selectedTaskIds).toEqual(["t3"]);
      },
    );
```

How can I resolve this? If you propose a fix, please make it concise.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[greptile-apps]

The schema uses positive() (≥ 1), but the context menu label renders "Archive ${taskCount} tasks" — if taskCount is 1 that reads "Archive 1 tasks". Changing to .min(2) also makes the schema self-document that this endpoint is intentionally bulk-only.

Suggested change

	export const bulkTaskContextMenuInput = z.object({
	taskCount: z.number().int().positive(),
	});
	export const bulkTaskContextMenuInput = z.object({
	taskCount: z.number().int().min(2),
	});

Prompt To Fix With AI

This is a comment left during a code review.
Path: apps/code/src/main/services/context-menu/schemas.ts
Line: 13-15

Comment:
The schema uses `positive()` (≥ 1), but the context menu label renders `"Archive ${taskCount} tasks"` — if `taskCount` is 1 that reads "Archive 1 tasks". Changing to `.min(2)` also makes the schema self-document that this endpoint is intentionally bulk-only.

```suggestion
export const bulkTaskContextMenuInput = z.object({
  taskCount: z.number().int().min(2),
});
```

How can I resolve this? If you propose a fix, please make it concise.