SkillFoxx

transcript-exporter skill

A Doubao skill that gathers Douyin titles, captions, transcripts and engagement data into Markdown and JSON or a Feishu table

  • 79 ★ on the repository
  • Last commit
Skill

Extended access

Why this level

  • The skill opens platform pages in the agent browser and collects other people's public video and creator data
  • On the fallback path the video audio is downloaded and uploaded to your Feishu for transcription
All reasons and checks

yuanyuanleah/transcript-exporterAuthor:Yuanyuan Leah

Install

Works in

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/transcript-exporter

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The skill opens platform pages in the agent browser and collects other people's public video and creator data
  • On the fallback path the video audio is downloaded and uploaded to your Feishu for transcription

This is third-party code. Review the repository files before installing.

What it does

transcript-exporter is a skill for the Doubao agent that collects Douyin video data. It accepts a creator page, a single video link or a short share link and extracts the title, caption, speech transcript and engagement: likes, comments, reposts, publish date and video URL. The skill saves results locally as Markdown and JSON or writes them to a Feishu multi-dimensional table. It can analyze what it collected: compare topics by performance, the comment-to-like ratio and publishing time spread. Progress is stored so an interrupted run resumes without duplicates. If a transcript cannot be obtained, basic data stays with the reason and no text is invented.

Who it is for. Short-video creators and editors who study competitor and own content, and analysts who keep a video database in Feishu.

Good fit when

  • You study a creator's last ten videos and want titles, transcripts and metrics in one table
  • You need to save the text and metrics of a single video from a short in-app link
  • A run stopped halfway and must continue without duplicating saved records
  • A team keeps an idea database in a Feishu table and refills it weekly
  • You compare topics and publishing times in collected data to pick the next story

Not a fit when

  • The platform is not Douyin: the skill is built around its pages and behavior
  • The agent is not Doubao: the instruction relies on that environment's built-in browser and `web.fetch`
  • You need a full account crawl without pauses: the skill paces itself and stops on platform checks
  • The data is for bulk extraction about third parties without their consent

Example request

For this Douyin video produce the title, full transcript and metrics and save locally, with no Feishu write

Limitations

The skill is written for Doubao and its built-in browser, and the README and guide are in Chinese. Collection is paced: by default 4–6 seconds between actions and a ten-minute rest after 15 videos, with a shorter fast mode. A login, slider or captcha check stops the run and hands it to a person; the skill does not bypass it. After two checks in 15 minutes it waits, and after three in a day it stops and delivers the cache. Writing to Feishu needs `lark-cli` and your authorization. The fallback transcript path downloads audio and uploads it to your Feishu.

How to disable. Remove the `skills/transcript-exporter` folder from the installed Doubao skills. Local Markdown and JSON results are stored separately, and Feishu table rows are deleted by hand.

FAQ

How do I install the skill in Doubao?

Send Doubao the repository link and ask it to read `skills/transcript-exporter/SKILL.md` with `references` and install the skill. If that fails, install the whole folder following the repository guide.

Is a Feishu table required?

No. Local saving works without it, and writing to a table needs `lark-cli` and your authorization.

What does the skill do on a captcha check?

It stops browser actions, asks you to pass the check by hand and rereads the page afterwards. It does not try to bypass the check.

How many videos are processed by default?

For a creator page the five latest videos are taken. Set a different count or a top range in the request.

What if a transcript is unavailable?

The skill saves the title, caption and metrics and states the reason in the transcript field. It does not substitute made-up text.

Discussion

No questions or reviews yet.

Skills

With douyin-comments, an agent collects comments and replies from one Douyin video into a report and local source records.

SkillExtended access531

An agent works in Douyin through local Chrome: skills and a JSON CLI find posts, prep a publish and comment, with confirmation

SkillBroad access61

A skill that audits accounts and viral posts on Xiaohongshu, Douyin, Kuaishou, WeChat Channels and Bilibili, then drafts a copy.

SkillExtended access369

A skill that prepares a video, title and cover for five platforms and stops short of the final Publish button.

SkillBroad access276

14 skills and a CLI save videos, podcasts and articles to a local Markdown base and find quotes in it with a source link

SkillBroad access1.3K

The sau CLI and skills for Claude Code and Codex: publish videos, articles and images to Douyin, Bilibili, YouTube, Instagram and more

SkillBroad access16.5K

Comparison with similar tools

Cards that share the main search query, and how they differ.

ToolOfficialRepository starsMain difference
transcript-exporter (this card)no79
douyin-to-obsidianno206saves a processed video into Obsidian notes; here the export goes to Markdown, JSON and a Feishu table with metrics
Douyin Favorites to Knowledgeno104collects favorites into a knowledge base; here a creator page, one video and topic comparison

README in short

The README describes a phone quick start through Doubao: copy the repository link, switch the task mode and ask it to install the skill. Version 1.2.0 runs continuously and reports progress every ten records, while the transcript fallback and Feishu writing sit in collapsible sections.

README badge

Original SKILL.md

---
name: douyin-transcript-exporter
description: 在豆包工作中整理抖音主页、单条视频或分享短链的标题、文案、逐字稿和互动数据，保存为本地 Markdown + JSON，或写入飞书多维表格。用户要求采集、导出或继续整理抖音视频资料时使用。
---

# 抖音逐字稿导出

## 执行原则

在豆包工作中执行：主页和详情页使用内置浏览器，逐字稿可使用其内置 `web.fetch`。不要求用户先配置可选依赖；某条路径不可用时继续其他路径并交付已有结果。

默认连续完成采集、保存和摘要，不逐条询问“是否继续”。普通校验、重试、落盘和进度告知不构成审批。已有清晰指令和授权直接复用；用户禁止下载、上传或写表时遵守该限制。

## Step 0：确定来源、输出和参数

- **来源**：接受主页、单条视频或分享短链。缺少来源且无法从上下文确定时询问一次；短链先用浏览器打开取得真实地址。
- **输出**：已有表格链接直接使用；说“本地保存”直接走本地。完全未说明时只询问一次“保存本地，还是写入飞书？”；用户说“跳过引导 / 直接开始”但未提供表格时，简短说明将先保存本地并开始，不再追问表格。
- **数量**：主页默认最新 5 条，用户给定数量或 Top 区间时按其要求；单条链接只处理该条。
- **逐字稿和 AI 字段**：默认提取逐字稿并生成选题方向、主题总结；“只要基础数据”时跳过这三项。只说“不要逐字稿”时仍可根据标题和介绍生成 AI 字段，并注明依据。
- **节奏**：未指定时直接采用稳妥模式，不强制用户先选；可按要求改为快速模式或自定义间隔。模式参数只在下一节定义。

### 采集节奏与中断

| 模式 | 操作间隔 | 休息 |
|---|---|---|
| 稳妥（默认） | 4–6 秒随机 | 每完成 15 条休息 10 分钟 |
| 快速（用户指定） | 2–3 秒随机 | 无固定批间长休息 |
| 自定义 | 使用用户给定间隔，如每条完成后休息 2–3 分钟 | 按用户要求 |

间隔是任务参数，不保证账号安全或固定耗时。等待期间保留进度并简短告知，不要求用户催办；不把长休息塞入短超时 cell。优先使用环境的等待/恢复能力；若环境无法继续等待，保存恢复点并如实说明需要恢复任务，不假称后台仍运行。

- 登录、滑块、点选或验证码出现时暂停相关浏览器操作，请用户接管手动完成；交回后重新读取页面并恢复，不尝试绕过。
- 对所有模式，15 分钟内连续两次验证先暂停 15 分钟；当日累计三次则保存进度、停止当日浏览器采集并交付缓存结果。已有缓存的整理和写入可以继续。计数按账号/日期保存在进度中，不因换批清零；这是本 Skill 的暂停策略，不是平台判定标准。
- 普通空白、加载失败或详情页错位只重试一次；仍失败记录原因并继续其他视频。重复 `video_id` 直接去重，不反复访问。
- 每完成 10 条或遇到需要用户接管的事件汇报进度；结束时报告不足 10 条的最后一批，不等待用户回复进度消息。

## Step 0.5：飞书准备（按需）

仅写入飞书时检测 `lark-cli` 是否为用户本人身份，并解析目标表。已有表格直接用，不重复介绍模板。只有用户需要建表时提供默认模板：可在用户指定的 Base 内新建数据表，或新建 Base；详见 [飞书写入](references/lark-write.md)。缺少目标且用户要跳过引导则先保存本地。

默认 11 字段：标题（text，主字段）、账号名称（text）、介绍（text）、逐字稿（text）、点赞/评论/转发（number）、发表日期（datetime）、视频链接（text）、选题方向（多选 select）、主题总结（text）。模板标签可按内容选择，已有表格的标签不自行新增。已有空主字段可重命名，不删除有数据的字段；不自动转换已有非空列的类型。

本地分支不检查飞书登录或字段。飞书权限缺失时保存本地结果，简短说明写入受阻；用户已指定写表但暂时失败时，仍保留该输出目标以便恢复，不能把本地缓存当作飞书已交付。

## Step 1：选定视频

Report a problem