Using Kairos Archive
Kairos Archive (this site and api.kairosarchive.net) is a stable-identifier registry for Sorcery: Contested Realm cards and printings. This page says what you may do with it, what belongs to whom, and what we ask of automated clients. No keys, no signup. The canonical text lives in the registry repository as docs/usage.md.
What Kairos Archive created, and how it is licensed
Everything the registry adds on top of the official data is free for any use, including commercial use, without permission: the identifiers and the guarantee that they never change; the record structure and histories; the derived fields and addresses; the schema; the documented data corrections; the code. Identifiers, structure and derived data are released under CC0 1.0; code under the MIT licence. Attribution is requested, not required: “Identifiers and structure: Kairos Archive (kairosarchive.net)”.
What belongs to Erik's Curiosa
Card names, rules text, typelines, flavour text, artist credits, set and product names, and every card image are © Erik's Curiosa and are used here with credit. The registry republishes the card data the official public API already serves, restructured for stability, and hosts card images itself as that API's own guidance asks. Both are displayed and served as part of the archive's function: keeping the record, matching printings to identifiers, and showing what an identifier refers to. This page says nothing about commercial use of Erik's Curiosa's content, either way, because that is not ours to grant or to withhold.
Practical requests
- Fetch in bulk, not object by object.
registry.jsonor theindex/lists; please do not crawl the object tree. - Cache what you fetch. Release roots are immutable; poll
versions.json(hourly is plenty) to learn of a new release. - Hotlinking images is allowed, with credit to Erik's Curiosa. Image addresses are permanent.
- Automated clients must send a
User-Agentnaming the project and a contact; requests with none are blocked at the edge. Browsers send one automatically. - Availability is best effort. Every release is mirrored on GitHub.
What we count
To know what the archive is used for, three things are counted, and nothing else. Page views and referrers, through Cloudflare Web Analytics, which sets no cookie and keeps no identifier. Clicks on links that leave the site: the host they go to and the page they were on. And requests to the query API, searches on the site, and the Discord bot: the route or command, how it ended, the client's software as a family name (“firefox”, “curl”), the country, the keys a query used (t e is:errata, never its words), how many results it found, and the time taken. Not stored: addresses, user ids, query text, message text, or the full User-Agent. The bot counts distinct servers through a keyed hash that cannot be turned back into a server id. When a card's name misses in the bot, the text that missed is kept, because that is how the matching improves. Requests to the data itself (the API host) are read from Cloudflare's sampled request analytics: paths, client software and countries, no addresses.
Kairos Archive may link printings to third-party sellers, including through affiliate programmes; such links are ordinary outbound links, disclosed where they appear.