
This takes podcast or video interview transcripts and turns them into structured writing material. It pulls out exclusive insights, counterintuitive claims, and quotable moments, then tags them by novelty and credibility. Beyond just extracting quotes, it analyzes the dialogue itself: which follow-up questions made the guest hesitate or pivot, where they dodged questions, where the conversation actually got deep versus stayed surface level. The output includes a chapter-by-chapter outline with key questions preserved and notes on what background info needs fact-checking. It's comprehensive to the point of being almost exhaustive, which is the right call when you're trying to turn a two-hour conversation into longform content.
$ npx -y skills add jimliu/illustrated-agent-skills --skill interview-analysis --agent claude-codeInstalls into .claude/skills of the current project.
将播客/视频访谈转化为结构化的写作素材包和提纲。
按主题分类,列出所有有价值的信息点,使用标签标注类型:
对每个关键论点,额外标注:
提取可直接引用的原话,筛选标准:
每条金句标注:原话内容、所属主题、使用建议(适合做标题/开头/论据等)
除金句外,单独提取以下写作素材:
分析访谈的对话结构,而非仅提取嘉宾观点:
列出需要搜索验证或补充的信息:
指出访谈中点到为止但值得追问的地方:
按访谈原始章节顺序生成写作提纲。每个章节包含:
[需补充: XXX的背景]Sponsors