概要
What is biliscribe?
biliscribe is an MCP (Model Context Protocol) server that extracts and formats content from Bilibili (B站) videos into structured text, optimized for processing and analysis by large language models (LLMs). It is intended for developers and users who need to convert video transcripts or metadata into machine-readable text for further LLM consumption.
How to use biliscribe?
Install the package globally using uvx mcp-server-biliscribe. Before running, ensure that BBDown and ffmpeg are available from your zsh shell. The server has been tested on macOS only. After installation, the server provides MCP tools that can be triggered by an MCP client to fetch and format Bilibili video content.
Key features of biliscribe
- Extracts Bilibili video content as structured text
- Output formatted for direct LLM input
- Requires BBDown and ffmpeg as external dependencies
- Published as a Python package via uvx
- Tested on macOS environment
Use cases of biliscribe
- Provide video transcripts to LLMs for summarization or analysis
- Automate conversion of Bilibili content into training datasets
- Integrate Bilibili video data into AI-powered workflows
FAQ from biliscribe
What external dependencies does biliscribe need?
You must have BBDown and ffmpeg callable from zsh. These are required to download and process the video content.
Which operating systems are supported?
The author has completed testing only on macOS. Running on other platforms may require additional configuration.
How do I install biliscribe?
Install using uvx mcp-server-biliscribe. No other installation methods are documented.
Does biliscribe store or cache any data?
The README does not mention any local storage or caching behavior. Data is processed on the fly from fetched Bilibili content.
What transport or authentication does biliscribe use?
The README does not specify transport or authentication details; it is an MCP server that likely uses standard MCP transport (stdio or SSE) via uvx. No authentication mechanism is described.
