via007

via007 / bilibili-rag

Public

B站收藏夹RAG知识库：收藏不吃灰，B 站收藏夹 → 语音转写 → 向量检索 → 对话问答

rag-bilibili-knowledge-base-llm-chroma-qwen
537
36
100% credibility
Found Feb 04, 2026 at 20 stars 27x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A web app that converts users' Bilibili video favorites into an AI-powered, searchable knowledge base for conversational querying.

How It Works

1
🔍 Discover BiliMind

You stumble upon this handy tool that promises to turn your saved Bilibili videos into a chatty personal knowledge base.

2
💻 Get it running

Follow a few easy steps on your computer to launch the web app right in your browser.

3
📱 Scan to connect

Use your phone's Bilibili app to scan a quick QR code and link your account securely.

4
📂 Browse your saves

See a list of all your favorite video collections, ready to explore.

5
⚡ Build your knowledge

Pick the folders you want, hit build, and watch it magically pull out key insights from your videos.

6
💬 Start asking away

Type natural questions in the chat, like 'What's the best beginner tip from my AI videos?'

🎉 Smart answers appear

Get helpful replies pulled straight from your videos, complete with links back to the exact sources, so your saves finally work for you.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 20 to 537 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is bilibili-rag?

This Python project transforms your Bilibili favorites—think hoarded tech talks, speeches, or courses like bilibili raghu thatha interviews or bilibili record of ragnarok season 3 clips—into a searchable RAG knowledge base. Scan QR login to pull folders, auto-transcribe audio via ASR (with ffmpeg fallback for restricted streams), embed chunks into a vector store, and query via chat UI that cites sources. Built on FastAPI backend and Next.js frontend, it delivers a local web app for building, updating, and conversing with your video collection.

Why is it gaining traction?

It solves the "favorites dust-up" problem for Bilibili power users: no more manual notes or lost gems amid bilibili raging fire uploads or raghu dakat bilibili marathons. Standout hooks include background syncing with progress tracking, semantic search across transcripts, and extras like cleaning invalid videos or auto-organizing default folders. At 323 stars, devs dig the end-to-end flow from login to RAG chat without API key hassles beyond DashScope.

Who should use this?

Bilibili addicts archiving learning videos—Chinese devs replaying Python tutorials, students recapping lectures, or anime fans querying bilibili ragna crimson plots via bilibili rags collections. Ideal for anyone with 50+ favorites folders wanting quick "summarize my ML playlist" Q&A, not scattered browsing.

Verdict

Grab it if you're deep in Bilibili content; solid docs, demo video, and test scripts make setup straightforward despite 1.0% credibility score and modest maturity (low test coverage, todo list like multi-part video support). Fork and extend for production—promising for personal RAG stacks.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.