A professional-grade skill for extracting Bilibili video subtitles and generating structured AI summaries using OpenAI-compatible APIs.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install videototext
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install videototext using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Videototext: Bilibili Subtitle Extractor & AI Summarizer?
Videototext is a specialized skill designed to bridge the gap between Bilibili video content and actionable text data. It provides a stable framework for fetching official subtitles, handling complex authentication requirements like SESSDATA and WBI signatures, and overcoming common rate-limiting challenges. By integrating this tool into your workflow, you gain access to high-quality transcripts that can be automatically summarized via any OpenAI-compatible LLM.
Built for the Openclaw Skills library, this tool is optimized for developers who need reliable extraction and synthesis. It includes sophisticated validation logic to ensure subtitle quality, such as title matching and duration coverage checks, making it an essential component for automated media analysis and research pipelines.
Videototext: Bilibili Subtitle Extractor & AI Summarizer Use Cases
- Automated content research and archival for Bilibili-based educational videos.
- Rapid summarization of long-form video tutorials or lectures into structured Chinese notes.
- Extracting transcripts for personal knowledge management systems like Obsidian or Notion.
- Batch processing video metadata and text content for large-scale data analysis.
How Videototext: Bilibili Subtitle Extractor & AI Summarizer Works
- The skill parses Bilibili URLs, including b23.tv short links, to extract the BV ID and page parameters.
- It initializes an authenticated session using provided cookies to access restricted or login-only subtitle tracks.
- The system queries Bilibili's WBI and web-interface APIs to locate the most relevant subtitle tracks.
- Raw JSON subtitle payloads are downloaded, cleaned, and merged into a continuous text format.
- The transcript is sent to a configured LLM endpoint, which generates a summary based on predefined prompts for different content lengths.
Videototext: Bilibili Subtitle Extractor & AI Summarizer Setup
To get started with this addition to your Openclaw Skills collection, follow these steps:
- Install the required dependencies in your environment:
pip install -r requirements-code.txt
- Copy the
.env.examplefile to.envin the skill root directory. - Configure your Bilibili credentials, specifically
SESSDATA, to enable authenticated subtitle extraction. - Set up your LLM configuration by defining
SUMMARY_LLM_BASE_URLandSUMMARY_LLM_API_KEY. - Zip the directory containing
SKILL.mdand thecode/folder for use with your AI agent.
Videototext: Bilibili Subtitle Extractor & AI Summarizer Data Schema & Taxonomy
The skill organizes extracted data and metadata using the following structure:
| Attribute | Description |
|---|---|
aid/bvid |
Unique identifiers for the Bilibili video source. |
subtitle_url |
The direct endpoint for the JSON subtitle track. |
transcript |
The concatenated text content generated from the subtitle payload. |
summary |
The AI-generated output, formatted with headers and bullet points. |
cookies |
Securely handled session data including SESSDATA and BILI_JCT. |
Videototext: Bilibili Subtitle Extractor & AI Summarizer Advanced Features
- WBI Player API support for accessing modern Bilibili subtitle tracks.
- Intelligent track prioritization prioritizing manual Chinese subtitles over AI-generated ones.
- Built-in request throttling and exponential backoff to handle 412 and 429 rate-limit errors.
- Subtitle validation engine that verifies text quality using title-match and coverage thresholds.
- Support for multi-part (multi-P) videos and short-link expansion.
SKILL.md
Loading
Related Openclaw Skills
Avalanche Analytics
A comprehensive intelligence tool for analyzing Avalanche subnets, DeFi protocols, and validator performance.
Ethereum L2 Analytics
A sophisticated intelligence tool for analyzing Ethereum Layer 2 scaling solutions, technical performance, and ecosystem liquidity.
Agent Rule Pilot
A centralized management tool for AI coding agent rules that synchronizes configurations across multiple platforms from a single Markdown file.
BSC Analytics
A comprehensive data and ecosystem analysis suite for the Binance Smart Chain powered by Openclaw Skills.
WeChat Article Reader
A specialized tool for extracting complete text, titles, and metadata from WeChat Public Account articles without requiring a login.
Sleep Story
A psychological sleep-aid tool that crafts 1000-2500 word healing bedtime stories using professional relaxation techniques.
Featured*
