An integrated strategy for archiving research and clips across media — YouTube video, BBC iPlayer, podcasts, and web, PDF and Kindle highlights via Glasp — into the federated wiki anarchive. Every capture becomes a wiki page with a low-res clip or quote in its assets folder: citable, forkable, searchable.
## The Pattern
Each channel follows the same pipeline: **capture** (download or export), **transcode** (a small web-native copy for the page, in the spirit of the Fair Use Clip), **page** (a wiki page with summary, source metadata and the media item), and **reindex**. The full-quality original is a temp file unless explicitly kept — the archive stores the reference and the working copy, not a mirror.
## YouTube
The most developed channel. yt-dlp — the actively maintained successor to youtube-dl — downloads the video, metadata, thumbnail and transcripts. The YouTube Archive Skill runs the whole pipeline: probe → title → download → sidecars → transcode → transcript-based summary → wiki page with chapters and an HTML5 video item → reindex. The Video FFmpeg Skill is the house transcode recipe — a 420px progressive-download copy that plays in a wiki video item.
## Podcasts
The podcast anarchive lives at podcast.anarchive.earth 
— see Downloading Audio for the capture technique and the p2p-archive intent in the spirit of Critical Commons. Podcast Index and per-show pages already exist there. What is missing is a `podcast-archive` skill mirroring youtube-archive: yt-dlp handles most podcast feeds and audio URLs too, so the same pipeline applies with an audio item instead of a video item.
## BBC iPlayer
The gap. `get_iplayer` is the established capture tool for BBC radio and TV; yt-dlp also covers many BBC URLs. Plan: an `iplayer-archive` skill on the same pattern, with the fair-use clip discipline mattering most here — short clips for citation and comment, not whole-programme mirrors.
## Highlights — Glasp
Reading is a media channel too. The Glasp Plan covers both directions: the official Glasp MCP connector for asking your highlight library inside Claude Code, and the REST API for a `glasp-to-fedwiki` sync that writes one wiki page per source document — articles, PDFs, YouTube and Kindle books — with highlights as quoted items. Glasp's category field routes each source to the right neighbourhood of the archive.
## Uploads
The channels above pull from platforms. Upload Media is the push side — our own voice notes, screencasts and footage arriving by form, Telegram bot or POST API, becoming pages through the same transcode–transcribe–page pipeline. Uploading Audio goes first; Uploading Video follows the ingress work.
## Infrastructure
The channels above now have a build plan of their own: the Anarchive Media Plan — one Media Plugin, one Outpost, and an IPFS pinning layer. This page stays the channel map; that page is the infrastructure plan.
## Next Steps
- build the three layers — plugin, Outpost, IPFS — per the Anarchive Media Plan phases - connect the Glasp MCP server and mint an API access token — Glasp Plan - build the `glasp-to-fedwiki` sync skill - generalise youtube-archive into a `podcast-archive` sibling - prototype `iplayer-archive` with get_iplayer - decide the archive's home neighbourhoods: this site for strategy and tooling, per-medium sites for the captures themselves
## See - Upload Media — the push side: uploads by form, bot and API
- Anarchive Media Plan — the infrastructure build plan: plugin, Outpost, IPFS - Glasp Plan — highlights into the archive - yt-dlp, youtube-dl — video capture tooling - YouTube Archive Skill and Video FFmpeg Skill — the pipeline automation - Downloading Audio, Fair Use Clip — the podcast side and the clip discipline - Hitchhikers Anarchive — the wider anarchive vision this serves
# Assets
media-plan