Require Cursor for every implementation unit

Why: Ensure implementation authorship is delegated per unit while the primary model remains responsible for planning and delivery management.

Evidence: The focused task-card generation regression went red then green, and git diff --check passed.

Impact: Every repository-writing implementation unit gets a fresh Cursor CLI run; unavailable or failed Cursor work is recorded as a blocker instead of being silently implemented by the primary model.

Refs: .task/feature-goal-first-delivery.md

tests/test_task.py

@@ -706,9 +706,22 @@ class TaskLifecycleTests(unittest.TestCase):
         card = (wt / ".task/feature-new.md").read_text()
         self.assertIn("make precheck", card)
         self.assertIn("交付模式：goal-first（目标、范围和交付终点一次确认；范围内连续推进）", card)
+        self.assertIn(
+            "实施执行：cursor-per-unit（每个产生仓库变更的实施单元启动一次新的 Cursor CLI 进程；"
+            "不得跨单元复用；主模型不直接撰写实施变更）",
+            card,
+        )
         self.assertIn("人工决策点：仅限目标/范围/验收变更、重大取舍、高后果操作或无法自动判定的业务验收", card)
         self.assertNotIn("authorized-continuous", card)
         self.assertIn("流程依据：仓库根目录 policy/delivery.md", card)
+        self.assertIn("## Cursor 实施收据\n", card)
+        cursor_receipts = card.split("## Cursor 实施收据\n", 1)[1].split("\n## ", 1)[0]
+        self.assertIn(
+            "| 单元 ID | 预期范围 | Cursor 运行 | 结果 | 阻塞原因 | 收据路径 |",
+            cursor_receipts,
+        )
+        self.assertIn("| 待定义 | — | NOT_RUN | — | — | — |", cursor_receipts)
+        self.assertIn("一单元一新进程；Cursor 不可用或失败时记录 blocker，不得交回主模型代写", cursor_receipts)
         for field in ("当前阶段", "当前增量", "方案引用", "实施计划引用", "测试计划引用"):
             self.assertIn(field + "：", card)
         self.assertIn("## 交付调度\n", card)