Bilibili Video Danmaku and Comment Analyzer

Problem

Audience feedback on Bilibili is scattered across danmaku, comments, and engagement signals. Exporting and reading raw data manually makes it hard to identify sentiment, discussion focus, and high-activity segments. This skill targets that workflow: given a Bilibili video URL or BV ID, it collects public API data and produces a structured Markdown report.

How It Works

The skill uses Bilibili public APIs and generally only needs a User-Agent request header, without login. The main steps are: resolve the bvid, fetch video metadata, pull XML danmaku using the oid or cid, page through comments, then run multi-dimensional analysis. It covers danmaku/comment sentiment, high-frequency keywords, danmaku timing distribution, engagement heat, and content quality, and finally exports a professional-style Markdown report. This helps identify which segments trigger concentrated discussion, which keywords recur, and whether feedback is positive, neutral, or mixed.

Boundaries

In an unauthenticated state, the comment API may return only a small subset, while full comments require login credentials. If the video has no danmaku, danmaku analysis is unavailable. Frequent requests may trigger 412/403 risk-control responses, so request pacing, delays, or a different User-Agent are recommended. The skill suits content review, light public-opinion monitoring, and video feedback analysis, but not large-scale crawling or exact full-comment statistics.