← Back to Skills Marketplace
danielcy

B站视频下载与解析

by danielcy · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ⚠ suspicious
382
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install bilibili-video-parser
Description
下载并解析B站视频。当需要执行分析视频内容等需要理解视频视觉信息时调用该技能
README (SKILL.md)

\r \r

Bilibili Video Parser\r

\r

使用方法\r

\r

注意\r

所有B站单条视频分析的需求,如果通过该技能无法执行,请直接通知用户错误内容,永远不要尝试使用web_fetch之类的工具去重试!\r \r

前置检查\r

\r

  1. 检查当前是否支持yt-dlp指令:\r
  2. 检查当前是否支持yt-dlp指令:\r
    • 如果不支持指令,可选择查看环境变量YT_DLP_PATH,如果该环境变量有值,则后续所有对yt-dlp的调用都替换为该路径\r \r
yt-dlp --version\r
```\r
\r
如果没有,直接终止并引导用户先至 https://github.com/yt-dlp/yt-dlp 进行安装\r
\r
2. 检查当前是否有`doubao-video-analyze`技能,如果没有,尝试先安装该技能\r
\r
### 视频链接获取\r
\r
0. 解析出视频的BV号,优先进行检查你的workspace目录下的./bilibili/videos目录下,是否已存在`{BV_ID}.mp4`,如果已存在,则直接进入`分析视频`步骤\r
1. 如果已知视频链接,则直接进入`下载视频`步骤\r
2. 如果知道视频的BV号,如`BV1s4ZLBAE22`,则直接按照如下规则拼接:\r
   `https://www.bilibili.com/video/BV1s4ZLBAE22`\r
3. 如果既没有视频链接,也没有BV号,则直接终止,并向用户询问视频链接或BV号\r
\r
### 下载视频\r
\r
0. 如果是已知路径的本地视频,则可以直接跳过该步骤,直接进入`分析视频`步骤\r
1. 需要将视频通过yt-dlp先下载至你的workspace目录下的./bilibili/videos目录下。如果找不到该目录,则先创建该目录\r
\r
```bash\r
yt-dlp -P {YOUR_WORKSPACE_DIR}/bilibili/videos -I 1 -o "{BV_ID}.mp4" {video_url}\r
```\r
\r
2. 该命令会打印出该文件的路径,你需要通过该行日志记住该路径\r
\r
```bash\r
[Merger] Merging formats into "xxx\bilibili\videos\{BV_ID}.mp4"\r
```\r
\r
3. 如果下载出错,直接终止任务,不要再继续进行其他尝试\r
\r
### 分析视频\r
\r
调用`doubao-video-analyze`技能分析该视频,直接使用该技能的结果
Usage Guidance
This skill appears to do what it says (use yt-dlp to download a Bilibili video and hand the file to a video-analyze skill). Before installing/using it: 1) Ensure yt-dlp is installed from the official source (https://github.com/yt-dlp/yt-dlp) and that YT_DLP_PATH (if set) points to a trusted binary on disk. 2) Be aware the skill will create and write files under your workspace (./bilibili/videos); confirm you are comfortable with that. 3) The skill may try to install or call the separate 'doubao-video-analyze' skill — review that skill before allowing automatic installation or autonomous agent actions. 4) YT_DLP_PATH is a path, not a secret; avoid setting it to anything unexpected (e.g., remote URLs or wrappers you don't trust). If you want tighter control, disable autonomous invocation or require manual approval before the agent installs or invokes other skills.
Capability Analysis
Type: OpenClaw Skill Name: bilibili-video-parser Version: 1.0.1 The skill bundle facilitates video downloading and analysis using the external utility `yt-dlp`. It is classified as suspicious because the instructions in `SKILL.md` direct the agent to construct shell commands using potentially unsanitized user input (`{video_url}` and `{BV_ID}`), which presents a significant risk of shell injection. Additionally, the instruction for the agent to autonomously attempt to install another skill (`doubao-video-analyze`) if missing could be leveraged to chain unauthorized software installations.
Capability Assessment
Purpose & Capability
Name/description match the declared runtime actions: it requires yt-dlp and an optional YT_DLP_PATH override to download Bilibili videos, and then delegates analysis to a video-analyze skill. Requests for yt-dlp are appropriate for a video-download parser.
Instruction Scope
SKILL.md is instruction-only and stays within expected boundaries (check yt-dlp, download to ./bilibili/videos, call doubao-video-analyze). It does instruct the agent to install/require another skill ('doubao-video-analyze') if missing and to create files under the workspace; those cross-skill install instructions are a functional dependency the user should be aware of. The doc references a workspace path placeholder ({YOUR_WORKSPACE_DIR}) that isn't declared in requires.env.
Install Mechanism
No install spec and no downloads are declared; this is an instruction-only skill (lowest install risk).
Credentials
Only YT_DLP_PATH is requested (and yt-dlp binary existence). This is proportionate for selecting a binary path. It's unusual that YT_DLP_PATH is marked as the 'primary credential' — it's a path, not a secret — so the metadata labeling is a minor inconsistency but not a functional problem.
Persistence & Privilege
always is false and the skill does not request permanent/global privileges. It does instruct the agent to install or call another skill at runtime, which could change the agent's capabilities if performed autonomously; that's a functional dependency rather than elevated privilege.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install bilibili-video-parser
  3. After installation, invoke the skill by name or use /bilibili-video-parser
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Added environment variable YT_DLP_PATH as a requirement for yt-dlp, allowing custom binary paths. - Specified "primaryEnv" as YT_DLP_PATH in metadata for improved configuration clarity. - No functional logic changes to the skill's operation.
v1.0.0
- Initial release of bilibili-video-parser. - Allows downloading and parsing Bilibili videos using yt-dlp. - Enforces pre-checks for yt-dlp and the doubao-video-analyze skill before processing. - Provides clear, step-by-step handling of video links or BV numbers. - Specifies strict error handling; no fallback retries with tools like web_fetch. - Integrates with doubao-video-analyze for video analysis after download.
Metadata
Slug bilibili-video-parser
Version 1.0.1
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is B站视频下载与解析?

下载并解析B站视频。当需要执行分析视频内容等需要理解视频视觉信息时调用该技能. It is an AI Agent Skill for Claude Code / OpenClaw, with 382 downloads so far.

How do I install B站视频下载与解析?

Run "/install bilibili-video-parser" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is B站视频下载与解析 free?

Yes, B站视频下载与解析 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does B站视频下载与解析 support?

B站视频下载与解析 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created B站视频下载与解析?

It is built and maintained by danielcy (@danielcy); the current version is v1.0.1.

💬 Comments