RAKKANBot
If you found RAKKANBot in your server logs, this page is for you. RAKKAN tracks announcements about manga, manhwa, manhua and anime: release dates, adaptations, broadcast dates, delays and hiatuses. It records who said what and when, and links each fact back to where it was announced. It never republishes the works.
User-Agent: RAKKANBot/<version> (+https://rakkan.cc/crawler)
Every request names itself. The one exception, while it lasts, is set out below.
Add this to your robots.txt:
User-agent: RAKKANBot Disallow: /
It takes effect the next time robots.txt is read, within six hours. To limit it to part of your site, disallow only those paths. Answering with 403, 405 or 429 also works: the source is marked failing and reviewed by a person rather than retried harder.
- RSS and Atom feeds that sites publish for readers.
- Public API endpoints, under each API’s own rules (openBD, AniList, 4chan’s read-only API). Reddit and Discord only through their official APIs, with the credentials and permissions they require.
- The public web preview of Telegram channels (t.me/s/<channel>).
- /robots.txt, before anything else on a site.
It does not follow links out of a feed, log in, fill in forms, get past paywalls or bot checks, or download images, video or chapter files.
- At most one request every two seconds to any one host.
- A feed is typically read every 30 to 60 minutes; registries once or twice a day.
- Conditional requests (ETag, If-Modified-Since): an unchanged feed costs you a 304.
- robots.txt is re-read at least every six hours.
For an announcement: its text, its address and when it was published, so a fact can be traced back to you. Items that name nothing RAKKAN tracks are not stored. For community sources that circulate leaked material, only the structured fact is kept (“no chapter next week”), never the text, a link or an image. To have a page removed, file a takedown notice.
robots.txt is the fastest and most reliable way to change what the crawler does. For anything else, the notice form reaches RAKKAN’s staff directly.
Temporary exception · until 2026-10-21
A few sites refuse RAKKANBot by name while serving the same public feed to any other client, and allow it in their own robots.txt. While we ask their owners for access, those feeds are read once more after a refusal under the HTTP library’s own name, python-httpx, at the same rate and never past a bot check: natalie.mu. A RAKKANBot rule in your robots.txt still stops both requests. This ends on 2026-10-21, or sooner if you ask.
Last updated 2026-09-23