About this skill

Problem

Bilibili comments are spread across videos, anime episodes, and posts. Manually paging through comments to find a keyword, regex pattern, or a specific user is slow, and the results are hard to reuse. For content review, public sentiment observation, or case retrospectives, you often need matched comments extracted into one place, plus a quick view of high-rated items and frequent users.

How it works

The skill focuses on Bilibili comment search and archiving:
- Supports bvid, epid, and dyid targets, and can parse a full Bilibili URL
- Supports keyword, regex matching, a specific uid, heat/time sorting, and image-only comments
- Supports searchSubReply for nested replies, reducing missed matches
- Produces a summary after search and writes JSON files to local results/
- Includes QR-code login and checkAuth; basic search can run without login

The typical flow is to extract the target and parameters, prefer the streaming search API, consume progress, match, summary, and complete events, then persist the results.

Boundaries

Results depend on Bilibili APIs, cookie validity, and risk-control behavior. The default request interval is about 500ms; do not lower it casually. MAX_PAGES controls scan depth, MAX_LIMIT controls how many matches are collected, and page only paginates displayed results. Enabling nested-reply search increases request volume, and popular videos cannot always be fully traversed. Cookies are stored locally, and search results are written only to the local directory.

Screenshots

Use Cases

Review a Bilibili tutorial video by extracting comments mentioning errors or steps and exporting top-rated matches.
Monitor an anime episode thread by regex-matching version or year mentions to surface key discussion points.
Compare posts from a specific UID across Bilibili posts, filter image-only comments, and save JSON results.
Sample public sentiment by paginating hot or time-ordered comments, collecting matches, and generating a summary.

Best For

Content operations staff reviewing Bilibili videos, who need to compile top-rated, keyword, and image-based comment feedback.
Analysts monitoring public sentiment, who need fast keyword or regex searches across videos, episodes, and posts.
Automation engineers, who need Agent-callable comment search, login checks, and local result storage.
Product managers investigating user feedback, who need to locate a UID or regex pattern in nested replies.

Related Skills

Antifragile Decision Framework

@user_15292d5a/yjkj-antifragile

Turns Taleb's antifragility, barbell strategy, and via negativa into executable prompts for risk analysis in career, investing, health, and other domains.

Knowledge Management 2026.08.30 ↓ 440