What is this?
Amane is a personal video library manager for the AI era: it watches your media folders, automatically scrapes metadata from multiple sources, and lets you browse the result through a web poster wall with an AI assistant for search and organization.
Is this a desktop app or a server I host myself?
Both — it ships as a native desktop app for macOS and Windows you can download and run directly, and as a Docker image for self-hosted server deployments.
How do I get it running?
Download the ready-to-use desktop build for macOS or Windows, or deploy the Docker image if you want it running as a server; full instructions are in the user documentation.
Are the community integrations officially supported?
No — community-built tools and plugins are maintained by third parties, and Amane only guarantees its public API; you have to judge their safety and functionality yourself.
Where does my scraped library data actually live?
Amane is local-first: all fetched metadata is persisted on your own machine or server, building a personal film repository instead of relying on a remote service.
What can the AI assistant actually do?
It lets you search your library with natural language, batch-organize files, and trigger metadata scraping jobs directly through chat.
How does the multi-source metadata scraping work?
Amane pulls metadata from several data sources at once and picks the best value field by field, instead of trusting a single source.
What does the image enhancement feature do?
It includes a built-in super-resolution tool that upscales and cleans up low-resolution poster images.
Are MovieDataCapture and AVDC the same project?
No, they are listed as related/prior projects (yoshiko2/MovieDataCapture and moyy996/AVDC) with similar scraping goals; Amane is a separate, newer implementation.
What are similar projects?
stash, TikTokDownloader, capcut-mate, autoclip, Kavita, ALH-Pro, Drift, dance-video-to-prompt
Which topics is it in?
media-server, metadata, super-resolution, scraper, ai-assistant, video-processing, image-processing, video, dashboard, local-first, gplv3, desktop-app