AGENT SKILL

account-video-downloader

暂无简介

使用场景图像与多媒体
License 许可
Apache-2.0
最近更新
2026-11-14
Stars
★ 1.6k
Forks
491

02 资源文件 9

03 SKILL.md 正文

Note: {SKILL_DIR} in the commands below means the absolute path of the directory containing this SKILL.md — resolve it before execution.

Channel attribution: replace claude in the REDFOX_CHANNEL=claude prefix below with the AI client you are currently running in: claude (Claude Code), codex (OpenAI Codex), cursor (Cursor), gemini (Gemini CLI). If the client is unknown, omit the prefix entirely — the script auto-detects the client from environment markers and falls back to github.

Multi-Platform Account Video Extractor

Platform + account -> fetch works list -> resolve download links -> one-click download to local


Overview

A batch downloader for account homepage videos across multiple platforms. It supports Douyin, Kuaishou, Bilibili, and YouTube: provide the platform name and account identifier, and it automatically fetches the account's recent works, resolves the direct download link for each video/image post, and supports one-click batch downloading to your machine.


Features

Feature Description
? Work fetching Get recent homepage works by account (title, engagement data, work link)
? Link resolution Call the resolve API per work to get the direct video/image download link
? Batch download One-click download of all works (video + image posts) to the local output/ directory
? Data display Markdown table / JSON dual-format output with full engagement data
? Pagination Page through to view more works
? Date filtering Filter by work publish date
? Multi-platform One tool covering Douyin / Kuaishou / Bilibili / YouTube

Supported Platforms

--platform Platform Account identifier Example
douyin Douyin Douyin ID e.g. Fish688688
kuaishou Kuaishou Kuaishou ID (kwaiId) e.g. Fish688688
bilibili Bilibili Homepage URL (accountUrl) e.g. https://space.bilibili.com/123456
youtube YouTube Channel URL (channel) e.g. https://www.youtube.com/@channel

API

This skill calls the redfox.hk API. Each platform takes two steps:

Step Endpoint Description
1. Fetch works POST /story/api/{platform}/... Get the works list by account identifier
2. Resolve download POST /story/api/parseWork/videoDownload/{platform} Get the direct download link by work URL

Authentication

Get an API Key at RedfoxHub and set it as an environment variable:

# macOS / Linux
export REDFOX_API_KEY=ak_your_key

# Windows PowerShell
$env:REDFOX_API_KEY="ak_your_key"

Usage

CLI

# Douyin: basic usage
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform douyin --account "Fish688688"

# Kuaishou: basic usage
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform kuaishou --account "kwaiId"

# Bilibili: fetch works and resolve download links (no file download)
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform bilibili --account "https://space.bilibili.com/123456"

# YouTube: download videos to local
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform youtube --account "https://www.youtube.com/@channel" --download

# YouTube: specify a download directory
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform youtube --account "https://www.youtube.com/@channel" --download --output-dir ./my_videos

# Set the number of works (default 10, max 50)
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform bilibili --account "MID" --count 20 --download

# Page through to view more works
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform kuaishou --account "userID" --page 2

# Filter works by date range
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform bilibili --account "MID" --date-start 2026-07-01 --date-end 2026-07-31

# Multiple accounts (comma-separated)
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform bilibili --accounts "MID1,MID2" --download

# Combined: page 2 + date filter + download
REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform bilibili --account "MID" --page 2 --count 20 --date-start 2026-06-01 --download

Parameters

Parameter Description
--platform Target platform (required, see table above)
--account A single account identifier
--accounts Multiple account identifiers, comma-separated
--count Number of works to fetch (default 10, max 50)
--page Page number (default 1)
--date-start Start date YYYY-MM-DD
--date-end End date YYYY-MM-DD
--download Download video files to local
--output-dir Download directory (default output/)
--json Output in JSON format
--rate-limit Seconds between requests (default 1.0)

Dependencies

Dependency Install command
requests pip3 install requests

Account Identifier Requirements

Each platform has different requirements for the account identifier — you must provide the platform's unique identifier:

Platform Requirement URL support How to obtain
Bilibili Homepage URL (accountUrl) ✅ Supported Copy the personal-space page URL directly
YouTube Channel URL (channel) ✅ Supported Copy the channel page URL directly
Douyin Douyin ID (uniqueName) ❌ Not supported Douyin APP -> target profile -> the 'Douyin ID: xxx' field below the avatar
Kuaishou Account ID (kwaiId) ❌ Not supported Kuaishou APP -> target profile -> the ID shown below the nickname

⚠️ Douyin and Kuaishou do not support homepage links! If the user pastes a URL, you must prompt them for the unique identifier above.

Important: if the user only enters an account name without the unique identifier, you must prompt them for the exact unique identifier.

? How to find the Douyin ID: open the Douyin APP -> go to the target user's profile -> below the avatar and follower stats you'll see the 'Douyin ID: xxx' field. If the user isn't sure where it is, show an annotated screenshot (with the 'Douyin ID' location highlighted) to help them locate it.


Agent Integration Guide

Trigger words

  • homepage video download / batch video download / account video extraction
  • Douyin download / Douyin video extraction / download Douyin works
  • Bilibili video extraction / Bilibili video download
  • YouTube channel download / YouTube video extraction
  • "download the homepage videos of xxx"

Agent execution flow

Step 1: Confirm the user's intent, platform, and account identifier
  - Identify the target platform from the user's input (Douyin/Kuaishou/Bilibili/YouTube)
  - If the platform is unclear, ask: "Which platform do you want to download from?"
  - Confirm the account identifier format is correct; if the user gave a nickname, prompt for the unique ID

Step 2: Recognize any time range in the user's input
  - If the user mentions a work time range, recognize it and convert to --date-start / --date-end:
    - "7.1~7.20" -> --date-start 2026-07-01 --date-end 2026-07-20
    - "July 1 to July 20" -> --date-start 2026-07-01 --date-end 2026-07-20
    - "the last week" -> 7 days back from today
    - "last month" -> the 1st to the last day of the previous month
    - "works from July" -> --date-start 2026-07-01 --date-end 2026-07-31
  - Dates are always YYYY-MM-DD; the year defaults to the current year
  - If the user asks to "continue"/"next page", add --page N

Step 3: Call the script to fetch works + resolve download links
  REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform <platform> --account "<id>"

Step 4: Show results + ask about pagination
  - Show the Markdown table with clickable work links and download links
  - On failures: note "the video may have been deleted by its author; for data verification contact redfoxdata@proton.me"
  - If more pages exist: tell the user "more works are available — view the next page?" and page on request
  - Remind the user that extracting by time range is supported

Step 5: Ask whether the user wants a batch download
  - Use AskUserQuestion: "Batch-download the downloadable videos to your machine?"
  - Options: "Download to local" / "Just show the links"

Step 6: Run the download after the user confirms
  REDFOX_CHANNEL=claude python3 "{SKILL_DIR}/scripts/main.py" --platform <platform> --account "<id>" --download
  - Tell the user where the files were saved

Agent output spec

After resolution, output in the following format (natural-language style, aimed at ordinary users).

⛔ Mandatory rule: do not simplify the output

The Agent MUST render the full table returned by the script verbatim. The following are forbidden:

  • ❌ Do not drop or merge any column (Published, Work, Likes, Comments, Saves, Shares, Download — all are required)
  • ❌ Do not omit the pagination hint line
  • ❌ Do not omit the date-range hint ("? You can specify a publish-date range to extract…")
  • ❌ Do not omit the download prompt ("? Want to batch-download these X works to your machine?")
  • ❌ Do not replace the full table content with "..." or a summary
  • ✅ Every work must be shown line by line with its full info (including resource download links)

Output template

## ? {Platform} Video Download — @accountName (followers: X)

Currently on **page 1**, 10 works in total | more works available, pass `--page 2` to view the next page

| #   | Published | Work              | Likes | Comments | Saves | Shares | Download                                      |
| --- | --------- | ----------------- | ----- | -------- | ----- | ------ | --------------------------------------------- |
| 1   | 07-28     | [Work title](link)| 5.2k  | 67       | 689   | 5.8k   | [?Video](...) · [?Cover](...) · [?Audio](...) |
| 2   | 07-24     | [Work title 2](link)| 3.1k | 22      | 136   | 865    | [?Cover](...)                                |
| 3   | …         | …                 | …     | …        | …     | …      | …                                             |

**Total:** 10 works, 8 downloadable, 2 failed

> ⚠️ A failed video may have been deleted by its author. For data verification, contact the support email **redfoxdata@proton.me**.

> ? Want works from a specific time range? Just tell me the range, e.g. "7.1~7.20" or "the last week".

> ? Want to batch-download these X works to your machine? Just let me know.

The platform name is substituted automatically for the actual platform (Kuaishou / Bilibili / YouTube).


FAQ

Q: How do I get an API Key? A: Register at redfox.hk to get one.

Q: Are the downloaded videos watermarked? A: No. The direct video/image links returned by the API are watermark-free.

Q: Can image posts be downloaded? A: Yes. Image posts (slideshows, galleries) automatically download the first image as JPG/PNG/WebP.

Q: Why do I see "rate limit exceeded"? A: The API returns code=3108 when requests come too fast and trigger throttling. Increase the interval with --rate-limit 2.0.

Q: Which platforms are supported? A: Douyin, Kuaishou, Bilibili, and YouTube.

Q: Where do I find each platform's account identifier? A: Douyin needs the Douyin ID (e.g. JCLjiangchenglan); Kuaishou needs the account ID / kwaiId (below the account name); Bilibili needs the homepage link; YouTube needs the channel URL.

04 同仓库更多 Skills

hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-review-with-annotations

Create, inspect, reply to, update, or resolve situated Worktable annotations. Use when feedback belongs to a Doc section, block, text range, or HTML Doc, or when the user asks to address existing review feedback and coordinate a requested artifact edit.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-find-and-synthesize-context

Find, assess, and synthesize context from Worktable Docs and Records. Use when the user asks what a workspace contains, wants workspace-backed research, needs current facts reconciled across artifacts, or requests a cited synthesis without changing workspace state.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-update-docs

Create, edit, rename, or maintain narrative Worktable Docs. Use for plans, decisions, briefs, notes, research syntheses, and other durable prose when the user requests a new artifact or a safe update to an existing one.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-records

Organize repeated structured information as durable Worktable Records. Use for research datasets, inventories, directories, pipelines, trackers, and other collections whose items need identity, fields, filtering, links, validation, or individual updates.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-create-or-manage-interactive-html

Create or improve durable visual and interactive artifacts in Worktable. Use for dashboards, trackers, calculators, mini apps, slide presentations, and other requests that benefit from a polished HTML Doc.

★ 1.6k
hashgraph-online
hashgraph-online/awesome-codex-plugins

worktable-collaborate-threads

Read, answer, and hand off messages through durable Worktable threads with users or other authenticated participants. Use when the user asks to check addressed messages, reply to a participant, continue a thread, request another agent's help, wait briefly for a response, or watch for comments and messages sent to the agent.

★ 1.6k