Douyin Video Fetch

下载抖音视频到本地（无水印优先）。用于给后续视频分析/复刻提供原始素材，支持 URL 或 video_id 输入、批量列表输入与统一输出目录。

Security Scan

The skill's code matches its stated purpose (downloading Douyin videos) but the package omits dependency and install details and executes a headless browser (running remote JS) without making those requirements explicit — review and run in isolation before use.

Purpose & Capability

Name/description (download Douyin videos) align with the included script which uses Playwright and aiohttp to load pages, intercept network responses, extract media URLs, and save MP4s. However, the skill metadata claims no required binaries/env but the code requires Python packages (playwright, aiohttp) and a browser runtime — this mismatch is notable.

Instruction Scope

SKILL.md simply tells the agent to run the script with a URL/video_id or file; it does not request unrelated files or secrets. The script loads Douyin pages in a headless browser and executes/collects in-page JS/data (via page.evaluate and response interception), which means it will run remote site JavaScript in the agent environment — expected for scraping but a potential risk if you haven't reviewed the code fully.

Install Mechanism

There is no install spec. The script depends on Playwright (and its browser binaries) and aiohttp, but these are not declared. Playwright typically requires installing browser engines (or calling 'playwright install'); omission of installation/dependency guidance is an operational and security concern because users may run the script in environments without those safeguards or isolation.

Credentials

The skill declares no environment variables or credentials and the code does not appear to request unrelated secrets. Network access is required (to fetch pages/media) but that is proportional to the stated purpose.

Persistence & Privilege

The registry flags show always=false and the skill does not attempt to modify agent/global configuration. It runs on-demand and does not request persistent elevated privileges.

What to consider before installing

This skill appears to do what it claims (download Douyin videos) but you should not install/run it blindly. Before using: 1) Inspect the full script to ensure there are no hidden network sinks or data exfiltration paths (the provided file was truncated in the review). 2) Install and run in an isolated environment (sandbox or container) because it launches a headless browser and executes page JS. 3) Ensure required dependencies are installed (Python, playwright, aiohttp) and run 'playwright install' to get browser binaries. 4) Consider legal/ToS implications of scraping/downloading content. If you want higher confidence, request the full unobfuscated source and an explicit dependency/install instruction from the publisher.

Overview

把抖音链接下载成可分析的本地 mp4。 这是“视频复刻”链路的素材入口层。

何时使用

• 你需要把目标视频落地到本地做拆解
• 你拿到的是 video_id，想直接下载
• 你要批量下载一组抖音视频做样本库

快速用法

单条下载：

python scripts/fetch_video.py "https://www.douyin.com/video/7599980362898427178"

用 video_id 下载：

python scripts/fetch_video.py 7599980362898427178

批量（每行一个 URL 或 video_id）：

python scripts/fetch_video.py --file input.txt --output-dir ./downloads/douyin

输出

• 默认输出目录：./downloads
• 文件名：<video_id>.mp4
• 终端会输出每条的成功/失败结果与落盘路径

备注

• 该技能只负责下载，不做ASR/镜头分析。
• 下载失败时建议先用 douyin-url-resolver 清洗输入链接。