[错误] L1 提取可以将第三方的名称分配给用户（提示中没有名称来源规则）

(Chinese summary) L1 prompts require "user (name)" and "I (AI) am with xxx (user identity) doing…", but **do not constrain the source of the name**: when a third party's name appears in L0 (user pasted a colleague's messages verbatim ("Mentor: pull the code and take a look") while the actual user's name rarely appears), the model treats it as the user; the `previousSceneName` continuity propagates the wrong name into the following batches (we observed 39 L1 records and their `scene_name` used a third party's name as the user on one day; L2/L3 with persona context had the attribution right, so this is L1-specific). It would be great to send a PR adding that rule to the prompt (with a small regression test) if the direction sounds right. An alternative would be to pass a user-identity hint from the L3 persona into the L1 prompt.