=== admin_ai_model ===
1 | 豆包(Doubao) | doubao | ep-20260605161826-r2b4m | status=1 | key=ark-17c6... | ep=https://ark.cn-beijing.volces.com/api/v3/responses
2 | 通义千问(Qwen) | qwen | qwen-plus | status=1 | key=EMPTY | ep=https://dashscope.aliyuncs.com/compatible-mode/v1/chat/compl
3 | 元宝/混元(Yuanbao) | yuanbao | hunyuan-lite | status=1 | key=EMPTY | ep=https://api.hunyuan.cloud.tencent.com/v1/chat/completions
4 | DeepSeek | deepseek | deepseek-chat | status=1 | key=sk-56706... | ep=https://api.deepseek.com/v1/chat/completions

=== scenes (lobster related) ===
quality_scoring #0 generator -> 豆包(Doubao) (doubao) status=1
quality_scoring #1 scorer -> 通义千问(Qwen) (qwen) status=1
quality_scoring #2 scorer -> DeepSeek (deepseek) status=1
semantic_analysis #0 generator -> 豆包(Doubao) (doubao) status=1
workflow_evolution #0 generator -> 豆包(Doubao) (doubao) status=1
workflow_llm #0 generator -> 豆包(Doubao) (doubao) status=1

=== all scenes ===
content_moderation
corpus_analysis
dynamic_adjustment
dynamic_node
event_decision
feedback_evolution
multi_turn_dialogue
quality_scoring
semantic_analysis
summary_generation
takeover_detection
template_evolution
tenant_learning
user_optimization
vector_embedding
workflow_evolution
workflow_fallback
workflow_generation
workflow_llm