fix(ssa): Fix 7 integration bugs and refactor frontend unified state management

Bug fixes:
- Fix garbled error messages in chat (TypeWriter rendering issue)
- Fix R engine NA crash in descriptive.R (defensive isTRUE/is.na checks)
- Fix intent misclassification for statistical significance queries
- Fix step 2 results not displayed (accept warning status alongside success)
- Fix incomplete R code download (only step 1 included)
- Fix multi-task state confusion (clicking old card shows new results)
- Add R engine and backend parameter logging for debugging

Refactor - Unified Record Architecture:
- Replace 12 global singleton fields with AnalysisRecord as single source of truth
- Remove isWorkflowMode branching across all components
- One Analysis = One Record = N Steps paradigm
- selectRecord only sets currentRecordId, all rendering derives from currentRecord
- Fix cross-hook-instance issue: executeWorkflow fallback to store currentRecordId

Updated files: ssaStore, useWorkflow, useAnalysis, SSAChatPane, SSAWorkspacePane,
SSACodeModal, WorkflowTimeline, QueryService, WorkflowExecutorService, descriptive.R

Tested: Manual integration test passed - multi-task switching, R code completeness
Co-authored-by: Cursor <cursoragent@cursor.com>

backend/scripts/seed-ssa-intent-prompt.ts

@@ -79,7 +79,15 @@ const SSA_INTENT_PROMPT = `你是一个临床统计分析意图理解引擎。
 {"goal":"regression","outcome_var":"Death","outcome_type":"binary","predictor_vars":["Age","BMI","Smoking","Stage"],"predictor_types":["continuous","continuous","binary","categorical"],"grouping_var":null,"design":"independent","confidence":0.8,"reasoning":"用户想分析影响死亡率的因素，Death是二分类结局，其余变量作为预测因素纳入logistic回归"}
 \`\`\`
 
-### 示例 4：模糊表达 — 需要追问
+### 示例 4：统计学意义/检验
+用户: "Yqol和bmi是否有统计学意义？"
+数据画像中有: Yqol [numeric], bmi [numeric], sex [categorical], age [numeric]
+输出:
+\`\`\`json
+{"goal":"correlation","outcome_var":"Yqol","outcome_type":"continuous","predictor_vars":["bmi"],"predictor_types":["continuous"],"grouping_var":null,"design":"independent","confidence":0.85,"reasoning":"用户想了解Yqol和bmi之间是否存在统计学显著关系，两者都是连续变量，适合相关分析或回归分析"}
+\`\`\`
+
+### 示例 5：模糊表达 — 需要追问
 用户: "帮我分析一下这份数据"
 数据画像中有: 10个变量
 输出:
@@ -87,7 +95,7 @@ const SSA_INTENT_PROMPT = `你是一个临床统计分析意图理解引擎。
 {"goal":"descriptive","outcome_var":null,"outcome_type":null,"predictor_vars":[],"predictor_types":[],"grouping_var":null,"design":"independent","confidence":0.35,"reasoning":"用户没有指定任何分析目标和变量，只能先做描述性统计，建议追问具体分析目的"}
 \`\`\`
 
-### 示例 5：队列研究
+### 示例 6：队列研究
 用户: "我想做一个完整的队列研究分析，看看新药对预后的影响"
 数据画像中有: Drug [categorical, 2个水平], Outcome [categorical, 2个水平: 0/1], Age [numeric], Gender [categorical], BMI [numeric], Comorbidity [categorical]
 输出:
@@ -127,7 +135,7 @@ async function main() {
         content: SSA_INTENT_PROMPT,
         model_config: { model: 'deepseek-v3', temperature: 0.3, maxTokens: 2048 },
         status: 'ACTIVE',
-        changelog: `Phase Q v1.0: 5 组 Few-Shot + Confidence Rubric 客观化`,
+        changelog: `Phase Q v1.1: 6 组 Few-Shot (增加统计学意义示例) + Confidence Rubric 客观化`,
         created_by: 'system-seed',
       }
     });

backend/src/modules/ssa/services/QueryService.ts

@@ -411,7 +411,9 @@ export class QueryService {
 
     if (query.includes('比较') || query.includes('差异') || query.includes('不同') || query.includes('有没有效')) {
       goal = 'comparison';
-    } else if (query.includes('相关') || query.includes('关系') || query.includes('关联')) {
+    } else if (query.includes('相关') || query.includes('关系') || query.includes('关联')
+      || query.includes('统计学意义') || query.includes('显著') || query.includes('检验')
+      || query.includes('p值') || query.includes('有无差别')) {
       goal = 'correlation';
     } else if (query.includes('影响') || query.includes('因素') || query.includes('预测') || query.includes('回归')) {
       goal = 'regression';

backend/src/modules/ssa/services/WorkflowExecutorService.ts

@@ -395,6 +395,11 @@ export class WorkflowExecutorService extends EventEmitter {
       });
 
       // 调用 R 服务
+      logger.info('[SSA:Executor] Calling R service', {
+        step: step.stepOrder,
+        toolCode: step.toolCode,
+        inputParams: step.inputParams,
+      });
       const response = await this.rClient.post(`/api/v1/skills/${step.toolCode}`, {
         data_source: dataSource,
         params: step.inputParams,
@@ -410,6 +415,16 @@ export class WorkflowExecutorService extends EventEmitter {
       if (response.data.status === 'error' || response.data.status === 'blocked') {
         const rMsg = response.data.message || '执行失败';
         const classified = classifyRError(rMsg);
+
+        logger.warn('[SSA:Executor] R tool returned error', {
+          step: step.stepOrder,
+          toolCode: step.toolCode,
+          rMessage: rMsg,
+          rErrorCode: response.data.error_code,
+          rUserHint: response.data.user_hint,
+          classifiedCode: classified.code,
+        });
+
         return {
           stepOrder: step.stepOrder,
           toolCode: step.toolCode,

backend/src/modules/ssa/types/query.types.ts

@@ -74,13 +74,47 @@ export interface PrunedProfile {
 // 2. LLM 原始输出的 Zod Schema（静态版本）
 // ────────────────────────────────────────────
 
+const VALID_VAR_TYPES = ['continuous', 'binary', 'categorical', 'ordinal', 'datetime'] as const;
+
+const VAR_TYPE_ALIAS: Record<string, VariableType> = {
+  numeric: 'continuous',
+  integer: 'continuous',
+  int: 'continuous',
+  float: 'continuous',
+  double: 'continuous',
+  number: 'continuous',
+  real: 'continuous',
+  factor: 'categorical',
+  string: 'categorical',
+  text: 'categorical',
+  character: 'categorical',
+  char: 'categorical',
+  nominal: 'categorical',
+  boolean: 'binary',
+  bool: 'binary',
+  logical: 'binary',
+  dichotomous: 'binary',
+  date: 'datetime',
+  time: 'datetime',
+  timestamp: 'datetime',
+};
+
+function normalizeVarType(val: unknown): VariableType {
+  if (typeof val !== 'string') return 'continuous';
+  const lower = val.toLowerCase().trim();
+  if ((VALID_VAR_TYPES as readonly string[]).includes(lower)) return lower as VariableType;
+  return VAR_TYPE_ALIAS[lower] ?? 'continuous';
+}
+
+const varTypeSchema = z.preprocess(normalizeVarType, z.enum(VALID_VAR_TYPES));
+
 /** LLM 直接输出的 JSON 结构（Zod 校验用） */
 export const LLMIntentOutputSchema = z.object({
   goal: z.enum(['comparison', 'correlation', 'regression', 'descriptive', 'cohort_study']),
   outcome_var: z.string().nullable().default(null),
-  outcome_type: z.enum(['continuous', 'binary', 'categorical', 'ordinal', 'datetime']).nullable().default(null),
+  outcome_type: varTypeSchema.nullable().default(null),
   predictor_vars: z.array(z.string()).default([]),
-  predictor_types: z.array(z.enum(['continuous', 'binary', 'categorical', 'ordinal', 'datetime'])).default([]),
+  predictor_types: z.array(varTypeSchema).default([]),
   grouping_var: z.string().nullable().default(null),
   design: z.enum(['independent', 'paired', 'longitudinal', 'cross_sectional']).default('independent'),
   confidence: z.number().min(0).max(1).default(0.5),