Repository navigation
判断标准弹窗里按收藏夹选 AI 平台、开启思考和联网核实;AI 平台加备注名 - #482
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
分拣台的 AI 平台原来跟着侧边栏选的那个走,「开启思考」是全局开关,联网搜索没有。现在这三项都在「判断标准」弹窗里按收藏夹设:左边写标准,右边选 AI 平台、开启思考、联网核实。平台不支持的开关灰掉并写原因。没在弹窗里保存过的收藏夹和以前一样。「所有收藏夹」当作一个收藏夹,有自己的一份;它自己没写判断标准时,每个视频按所在收藏夹的标准判断。收藏夹设置里的「开启思考」一行去掉了。
联网核实只在字幕细看时生效,只对 DeepSeek 的 deepseek-flash 发(请求里加
deepseek_web_search)。打开后,没联网做的细看算旧结果,可以「按新标准重新细看」;换平台或开关思考不算。被平台内容审核拦掉的回复单独报错。AI 平台每行加可选的「备注名」,同一平台配多个 Key 时用它区分。侧边栏和判断标准弹窗里的平台选项写备注名,没填写「平台 · 模型」。
取舍和已知的行为:每个收藏夹的这三项存在本地,不进设置导出文件。细看结果按视频存、带着当时用的标准,所以在「所有收藏夹」用它自己的标准看过的视频,回到原收藏夹会出现「按新标准重新粗看 / 细看」,反过来也一样,不会自动重跑。收藏夹开着联网核实时,agent 经 MCP 写的细看结果也算「没联网做的」。agent 的
triage_settings读不到「所有收藏夹」那一份判断标准。验证:
node scripts/selftest.mjs50 项全过,node scripts/smoke.mjs通过。在临时配置的无头 Chrome 157 里加载扩展,用鼠标键盘事件走了真实的分拣台、设置页和侧边栏(B站 和 AI 平台的请求全部在本地拦截并伪造回复):原有功能 21 步、这个改动 16 步都通过,包括每个平台的开关可用性、保存后刷新还在、细看请求带搜索工具、「所有收藏夹」的回落、平台被删后的回落。另用真实的 DeepSeek Key 对 5 个视频各跑了不联网和联网两遍字幕细看(分支的后台代码,存储写入全部拦住):5 次联网回复都能解析,耗时 3.4–4.6 秒(不联网约 2 秒),每个视频搜 2–5 次,输入 token 从 0.4k–2.3k 涨到 12k–17k,2 个视频的结论变了。没测到:登录账号的真实数据、内容审核拦截的真实返回、开着思考再联网、AI 平台的授权流程。