feat: add AI学习助手 agent (KG+RAG ideal) and renshenguo feishu bot

- Add AI学习助手 agent creation script with all 39 tools, 3-layer KG+RAG memory
- Add renshenguo (人参果) feishu bot integration (app_service + ws_handler)
- Register renshenguo WS client in main.py startup
- Add RENSHENGUO_APP_ID / RENSHENGUO_APP_SECRET / RENSHENGUO_AGENT_ID config
- Reorganize docs from root into docs/ subdirectories
- Move startup scripts to scripts/startup/
- Various backend optimizations and tool improvements

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/services/workflow_engine.py

@@ -5525,7 +5525,6 @@ class WorkflowEngine:
 
                 # 调用 LLM 评判
                 try:
-                    from app.services.llm_service import llm_service
                     judge_prompt = (
                         "你是严格的内容质量评审专家。请根据以下标准对内容进行评分。\n\n"
                         f"【评判标准】\n{criteria}\n\n"
@@ -5700,9 +5699,13 @@ class WorkflowEngine:
                 can_execute = True
                 incoming_edges = [e for e in active_edges if e["target"] == node_id]
                 if not incoming_edges:
-                    if node_id not in [n["id"] for n in self.nodes.values() if n.get("type") == "start"]:
-                        logger.debug(f"[rjb] 节点 {node_id} 没有入边，跳过执行")
-                        continue
+                    # 节点无入边：必须是 start 类型，或整个工作流中没有 start 节点才可作为起点
+                    is_start_node = node_id in [n["id"] for n in self.nodes.values() if n.get("type") == "start"]
+                    if not is_start_node:
+                        has_any_start = any(n.get("type") == "start" for n in self.nodes.values())
+                        if has_any_start:
+                            logger.debug(f"[rjb] 节点 {node_id} 没有入边，跳过执行")
+                            continue
                 else:
                     for edge in incoming_edges:
                         src = edge["source"]