DanielW-Skills

把本地视频或音频直接转成豆包语音识别字幕文件的 standalone skill，自带运行脚本和配置模板。用户只要填自己的豆包 ASR 认证信息，就可以从一个本地视频得到 `srt` 字幕，并在视频同级生成一个命名为 `视频原名 + 豆包语音识别中间文件` 的目录，里面放词级时间戳、字幕分组、原始返回和抽出的音频。用户提到豆包语音识别、剪映字幕、SRT、视频转文字、口播转字幕、本地文件转写、怎么开始使用、怎么配置、怎么拿 API Key 时使用

Extract YouTube transcripts, clean them into timestamp-per-line markdown, and generate podcast summary + takeaways + mindmap + SEO blog post outputs. Use when a user provides a YouTube URL and wants transcript/summary/mindmap/blog assets saved under podcasts/<video-title>.

Review past bug-fix postmortem reports stored under `.agents/postmortem` and surface likely relevant historical fixes. Use when the coding agent is debugging or implementing a bug fix and wants examples of similar root causes, fixes, touched files, or verification patterns before proceeding.