Diagnose OP/QBot conversation, workflow, Agent, API, RAG retrieval, knowledge base indexing, multimodal and file-dialog issues. Use when investigating OP后台, QBot, record_id, session_id, request_id, trace_id, cfg_version_id, reply_method, prompt_raw, knowledge_raw, workflow, RunNodes, Agent, handoff, tool_call, task_done, Galileo, Trace, 日志, API报错, 请求失败, QA未命中, 兜底, 样表或示例数据误答, chunk, segment, vector index, embedding, rerank, RRF, 文件解析, 图片问答, 图文检索, 实时文档问答. Start by classifying the app mode, then use OP structured data and Galileo runtime evidence to produce an evidence-based diagnosis card.
Diagnose OP/QBot conversation, workflow, Agent, API, RAG retrieval, knowledge base indexing, multimodal and file-dialog issues. Use when investigating OP后台, QBot, record_id, session_id, request_id, trace_id, cfg_version_id, reply_method, prompt_raw, knowledge_raw, workflow, RunNodes, Agent, handoff, tool_call, task_done, Galileo, Trace, 日志, API报错, 请求失败, QA未命中, 兜底, 样表或示例数据误答, chunk, segment, vector index, embedding, rerank, RRF, 文件解析, 图片问答, 图文检索, 实时文档问答. Start by classifying the app mode, then use OP structured data and Galileo runtime evidence to produce an evidence-based diagnosis card.