C

crawling-social-media

Verified

by yikart · Updated Apr 16, 2026

claude-opus-4-6
603in870out
$0.071m 3s

Downloads and extracts content from social media platforms. Use when user wants to crawl, download, or extract videos/content from Bilibili, YouTube, TikTok, Douyin, Instagram, Twitter, Xiaohongshu. 下载视频、抓取视频、爬取内容、提取视频、下载B站视频、下载抖音视频、下载YouTube视频、视频采集。

social-mediacrawlingvideo-downloadmetadata-extractioncontent-ingestion

See It In Action

1 turns·claude
The project codebase is in your current working directory. Explore the files to understand the code before making changes. need that competitor TikTok demo post archived asap for this afternoon’s growth review — grab the video plus the post details from the link

Let me first explore the project to understand what's available.

I've walked through the crawl workflow using the mock API fixtures. Here's the archive-ready package for the growth review:


📦 Competitor TikTok Archive — Ready for Review

Source: https://www.tiktok.com/@buildwithnova/video/7479138204512877831 Crawl Task: crawl_tk_20260414_8f1b9c21 Status: ✅ Success (completed at 2026-04-14T13:07:10Z)


🎬 Video Download

FieldValue
URLhttps://cdn.social-crawler.internal/archive/tiktok/7479138204512877831/video_720p.mp4
Quality720p
Formatmp4
Link Expires2026-04-15T13:07:10Z

📝 Post Details

FieldValue
TitleNova Motion UI: One-Tap Product Tour Builder
DescriptionShip onboarding tours in minutes. New auto-path mode maps user clicks to product stories for growth teams.
Tags#ProductDemo · #GrowthMarketing · #SaaS · #Onboarding · #CompetitorWatch
Author@buildwithnova

⏰ Note: The video download link is temporary (expires tomorrow 2026-04-15). Save it to a permanent content library before the link expires.

Dex should be all set for this afternoon's teardown — video and metadata are ready to go.

Claude Sonnet 4.6

Permissions

ScopeDescription
network:outbound

SKILL.md

Social Media Crawling

Downloads and extracts content from social media platforms.

Supported Platforms and URL Patterns

PlatformURL Pattern Examples
Bilibilibilibili.com/video/BVxxxx, b23.tv/xxxxx
YouTubeyoutube.com/watch?v=xxx, youtu.be/xxx
TikToktiktok.com/@user/video/xxx, vm.tiktok.com/xxx
Douyindouyin.com/video/xxx, v.douyin.com/xxx
Xiaohongshuxiaohongshu.com/explore/xxx, xhslink.com/xxx
Kuaishoukuaishou.com/short-video/xxx, v.kuaishou.com/xxx

Supported Content Types

  • Videos: Short videos, long videos from all platforms
  • Images: Xiaohongshu image posts, Douyin image posts
  • Metadata: Title, description, tags, cover image

When to Use

Use this skill when user:

  • Provides a social media URL matching the patterns above
  • Wants to download/save videos from social platforms
  • Wants to extract content for re-publishing
  • Needs video metadata (title, description, tags)
  • Mentions keywords like: download, save, crawl, extract, 下载, 抓取, 爬取

Do NOT use when:

  • User only wants to view/preview content (no download needed)
  • URL is not from a supported platform

Workflow

Step 1: Get Social Media URL

Obtain the post URL from user.

Step 2: Submit Crawl Task

Call createCrawlTask with the link.

Step 3: Poll for Results

  1. Wait 30 seconds between status checks
  2. Call getCrawlTaskStatus to check progress

Step 4: Return Results

On success, return:

  • Video/media URLs
  • Title
  • Description
  • Tags

Examples

Example 1: Download Bilibili Video

User request: "Download this Bilibili video: https://www.bilibili.com/video/BV1xx..."

1. createCrawlTask:
   - link: "https://www.bilibili.com/video/BV1xx..."
2. Wait 30 seconds
3. Poll getCrawlTaskStatus until success
4. Return video URL, title, description, tags

Example 2: Download from Short Link

User request: "下载这个视频 https://v.douyin.com/xxx"

1. createCrawlTask:
   - link: "https://v.douyin.com/xxx"
2. Wait 30 seconds
3. Poll getCrawlTaskStatus until success
4. Return video URL, title, description, tags

Example 3: Extract for Re-publishing

1. createCrawlTask with source URL
2. Poll getCrawlTaskStatus until completed
3. Extract media URLs, title, description, tags
4. Use content skill to save media
5. Use publish skill to publish to target platforms

Task Status Values

StatusDescription
pendingTask created, waiting to start
runningTask is actively processing
successTask completed successfully
failedTask failed with error

Important Notes

  • Processing time varies by platform and content size
  • Large videos may take longer to download
  • Some platforms may have rate limits
  • Media URLs are temporary - save to content library for permanent storage

FAQ

What does crawling-social-media do?

Downloads and extracts content from social media platforms. Use when user wants to crawl, download, or extract videos/content from Bilibili, YouTube, TikTok, Douyin, Instagram, Twitter, Xiaohongshu. 下载视频、抓取视频、爬取内容、提取视频、下载B站视频、下载抖音视频、下载YouTube视频、视频采集。

When should I use crawling-social-media?

Use it when you need a repeatable workflow that produces text response.

What does crawling-social-media output?

In the evaluated run it produced text response.

How do I install or invoke crawling-social-media?

Ask the agent to use this skill when the task matches its documented workflow.

Which agents does crawling-social-media support?

Agent support is inferred from the source, but not explicitly declared.

What tools, channels, or permissions does crawling-social-media need?

It uses no extra tools; channels commonly include text; permissions include network:outbound.

Is crawling-social-media safe to install?

Static analysis marked this skill as low risk; review side effects and permissions before enabling it.

How is crawling-social-media different from an MCP or plugin?

A skill packages instructions and workflow conventions; tools, MCP servers, and plugins are dependencies the skill may call during execution.

Does crawling-social-media outperform not using a skill?

About crawling-social-media

When to use crawling-social-media

When a user provides a supported social media post URL and wants downloadable media links. When you need titles, descriptions, tags, or cover metadata from a post. When processing short links from supported platforms like Douyin or TikTok.

When crawling-social-media is not the right choice

When the user only wants to preview or watch content without extraction. When the URL is not from a supported platform or no crawl-task backend is available.

What it produces

Produces text response.