feat(ssa): Complete QPER architecture - Query, Planner, Execute, Reflection layers

Implement the full QPER intelligent analysis pipeline:

- Phase E+: Block-based standardization for all 7 R tools, DynamicReport renderer, Word export enhancement

- Phase Q: LLM intent parsing with dynamic Zod validation against real column names, ClarificationCard component, DataProfile is_id_like tagging

- Phase P: ConfigLoader with Zod schema validation and hot-reload API, DecisionTableService (4-dimension matching), FlowTemplateService with EPV protection, PlannedTrace audit output

- Phase R: ReflectionService with statistical slot injection, sensitivity analysis conflict rules, ConclusionReport with section reveal animation, conclusion caching API, graceful R error classification

End-to-end test: 40/40 passed across two complete analysis scenarios.

Co-authored-by: Cursor <cursoragent@cursor.com>

r-statistics-service/tools/mann_whitney.R

@@ -173,6 +173,62 @@ ggplot(df, aes(x = .data[[group_var]], y = .data[[value_var]])) +
   labs(title = paste("Distribution of", value_var, "by", group_var))
 ')
   
+  # ===== 构建 report_blocks =====
+  log_add("构建 report_blocks")
+  blocks <- list()
+
+  # Block 1: 样本概况（两组 n, median, IQR）
+  g1_label <- as.character(groups[1])
+  g2_label <- as.character(groups[2])
+  blocks[[length(blocks) + 1]] <- make_kv_block(
+    title = "样本概况",
+    items = list(
+      list(key = paste0(g1_label, " (n, Median, IQR)"),
+           value = paste0("n=", n1, ", ", round(median(g1_vals), 3), ", ", round(IQR(g1_vals), 3))),
+      list(key = paste0(g2_label, " (n, Median, IQR)"),
+           value = paste0("n=", n2, ", ", round(median(g2_vals), 3), ", ", round(IQR(g2_vals), 3)))
+    )
+  )
+
+  # Block 2: 检验结果（U 统计量, Z 值, P 值, 效应量 r）
+  blocks[[length(blocks) + 1]] <- make_table_block(
+    title = "Mann-Whitney U 检验结果",
+    headers = c("U 统计量", "Z 值", "P 值", "效应量 r", "效应量解释"),
+    rows = list(
+      list(
+        round(as.numeric(U), 4),
+        round(z_value, 4),
+        format_p_value(result$p.value),
+        round(effect_r, 4),
+        effect_interpretation
+      )
+    ),
+    footnote = "Wilcoxon rank sum test with continuity correction"
+  )
+
+  # Block 3: 箱线图（如果 plot_base64 不为 NULL）
+  if (!is.null(plot_base64)) {
+    blocks[[length(blocks) + 1]] <- make_image_block(
+      base64_data = plot_base64,
+      title = paste0(value_var, " by ", group_var),
+      alt = paste("箱线图：", value_var, "按", group_var, "分组")
+    )
+  }
+
+  # Block 4: 结论摘要
+  sig <- if (result$p.value < 0.05) "存在统计学显著差异" else "差异无统计学意义"
+  blocks[[length(blocks) + 1]] <- make_markdown_block(
+    title = "结果摘要",
+    content = paste0(
+      "两组 **", value_var, "** 的比较（Mann-Whitney U 检验）：",
+      "U = ", round(as.numeric(U), 2),
+      "，Z = ", round(z_value, 3),
+      "，P ", format_p_value(result$p.value),
+      "，效应量 r = ", round(effect_r, 3), "（", effect_interpretation, "）。",
+      "两组间", sig, "。"
+    )
+  )
+
   # ===== 返回结果 =====
   log_add("分析完成")
   
@@ -209,6 +265,7 @@ ggplot(df, aes(x = .data[[group_var]], y = .data[[value_var]])) +
         )
       )
     ),
+    report_blocks = blocks,
     plots = if (!is.null(plot_base64)) list(plot_base64) else list(),
     trace_log = logs,
     reproducible_code = as.character(reproducible_code)