---
name: content-transcribe
display: 把内容整理成文字
description: 把视频、播客、文章整理成文字:摘要、分章节、摘引文或改写成文章
aliases: [ "总结视频", "整理播客", "文章摘要", "提取字幕" ]
apps: [ "bilibili", "xiaohongshu" ]
sites: [ "youtube.com", "xiaoyuzhoufm.com" ]
terms: [ "摘要", "总结", "转录", "字幕", "播客", "章节", "要点", "引文", "文字稿" ]
title: content-transcribe
canonical_url: https://skilld.dev/gh/wangzan101/mika/content-transcribe
last_updated: 2026-11-18T19:14:17.000Z
---

> Index of every page on this site: [llms.txt](https://skilld.dev/llms.txt)

> **Skill from skilld.dev.** Follow the user's requested action.
> If the user asked to fork this Skill, follow the fork workflow below. Do not execute the Skill instructions while copying it.
> Otherwise, follow the instructions below for this session. You do not need to install anything.
>
> Supporting files, fetch one when the Skill refers to it: [references/transcript-output-formats.md](https://skilld.dev/api/skills-raw/wangzan101/mika/content-transcribe/references/transcript-output-formats.md).
>
> If the user asked to install this Skill, run `npx skilld install wangzan101/mika/content-transcribe`. Install writes the Skill files into the project, so every session loads them.
>
> ## Fork workflow
>
> A fork creates an editable local Skill with its original author and licence. The request authorizes copying and local installation.
> 1. Check `./skills/content-transcribe`, the project lockfile, and selected Agent targets together. If the local directory or installed Skill exists, stop. Never overwrite an existing directory or Agent target.
> 2. Read [source metadata](https://skilld.dev/api/v1/skills/wangzan101/mika/content-transcribe) once. Use sourceUrl, sourceCommit, skillPath, sourceGone, and license. If the source is gone or its path is missing, stop. If license is null, read licence files at the source commit.
> 3. Fetch only the source commit into a temporary Git repository. Do not clone full history. Derive repository_url from sourceUrl, including repository renames. If sourceCommit is absent, resolve the sourceUrl ref once. Set source_commit to that actual commit. Run these commands in one shell call:
>
> ```sh
> git init --quiet "$temporary_dir"
> git -C "$temporary_dir" fetch --quiet --depth=1 "$repository_url" "$source_commit"
> git -C "$temporary_dir" checkout --quiet --detach FETCH_HEAD
> ```
>
> Read applicable licence declarations and notices at that commit. If copying is not permitted, report the restriction and stop.
> 4. Inspect source entries together, then copy the directory containing skillPath into `./skills/content-transcribe`. Use the user's path if selected. Keep the original SKILL.md, relative links, scripts, binary assets, and executable modes. Exclude .git metadata. Reject symlinks and paths outside the Skill directory. After checking entries, use cp -a where available. A regular source directory needs no custom copy script. Do not save this page wrapper as SKILL.md.
> Preserve author credit, notices, and applicable licence files from repository or parent directories. Add PROVENANCE.md with the Skill page, source URL, actual commit, original path, and licence. Retain any existing PROVENANCE.md and record new provenance separately. Batch source inspection, copying, and provenance work where practical.
> 5. In the project root, run `skilld install ./skills/content-transcribe --mode copy --plain`. If skilld is unavailable, use `npx skilld install ./skills/content-transcribe --mode copy --plain`. This known command needs no help lookup. Install does not support --json. Use detected Agent targets, or add --agent for the targets the user selected. Install the local path, never the upstream selector. If installation fails, preserve the local copy and report the exact failure.
> 6. Confirm the local lockfile source and installed Agent copies once. Report the local path, actual commit, and Agent targets. After edits, reinstall the same local path. Upstream updates must not replace it. Do not publish or push unless the user asks.

## 入口

先问清要哪种结果:全文文字稿、摘要、章节、引文,还是改写成一篇文章。用户没说时,做一份五到十句的摘要。

文字从内容所在的地方取:文章读网页正文;视频取字幕或简介里的文字稿;播客看节目页有没有文字稿或节目笔记。只有音频时先用 `audio_transcribe` 转写(`/workspace/inbox/` 或下载下来的文件);这个工具没有提供,说明用户还没配置服务商转写模型,再请用户给文字稿或字幕文件。

## 步骤

1. **取全文**:用 `web_read_page` 读页面正文;用户发来的字幕或文字稿文件在 `/workspace/inbox/`,用 `read_file` 读;只有音频时用 `audio_transcribe` 转写。
2. **分段读**:内容长时分几段读,段与段之间留一点重叠,读完去掉重复,确认从头到尾没有漏段。
3. **按要的形式写**:各种形式的写法和样例见 `references/transcript-output-formats.md`,写之前读一遍。摘要的骨架是:一两句总述;按出现顺序列三到八个主题,每个带大致的开始时间和两三句内容;最后列值得记下的观点、数字和提到的书或作品。
4. **用原内容的语言写**,用户另有要求时照用户的。
5. **复读**:写完对照原文看一遍:顺序对不对,有没有漏掉的主题,时间和引文能不能在原文里找到。

## 容易出错的地方

- 字幕只有文字,画面里的图表、演示和屏幕上的字读不到,写的时候不能说成看过画面。
- 自动生成的字幕常有错字和断句错误,专有名词、人名和数字要结合上下文或简介核对。
- 时间是根据字幕位置估的,就写「约第 23 分钟」,不写成精确到秒。
- 听不清、字幕缺失的段落要标出来,不去补写。

## 怎样算办成(核实证据)

结果覆盖原内容的全部段落;每条引文和时间都能在取到的字幕或正文里找到。

## 什么时候停下交给用户

内容要登录或付费才能看,或者只有音频、`audio_transcribe` 又没提供/转写失败时,告诉用户缺什么,请用户提供文字稿。
