# Almost all of this app is behind a login, so nothing below is secret -- a crawler # that ignores this file gets a 302 to the login form and learns nothing. What it costs # without this file is the request, the redirect, and a line in the log, on every page # of an account's history, for every crawler, forever. This is the cheapest way to say # "there is nothing here for you". User-agent: * # The public surface, and all of it. Five pages and the root, which redirects to the # first of them. Allow: /$ Allow: /welcome Allow: /about Allow: /docs Allow: /login Allow: /create-account # An account's own pages. /start is the empty-state guide and is behind a login too. Disallow: /start Disallow: /workouts Disallow: /exercises Disallow: /volume Disallow: /programs Disallow: /settings Disallow: /equipment # The OAuth endpoints, and the consent screen most of all: it is the one page in this # app where a single tap hands an API client the account, so it is the last page that # should ever be arrived at from a search result. Disallow: /connections Disallow: /authorize Disallow: /token Disallow: /oauth- # Discovery metadata. Public by design and fetched by OAuth clients, which do not read # this file; it is listed because a search index is not what it is public for. Disallow: /.well-known/ # The Sitemap line this file has been holding open, and #359 is the day it predicted: # documentation for the connector is indeed the first page it was waiting for. # # The original argument was that the public pages link to each other, so a sitemap would # tell a crawler only what it already has. /docs is linked too -- from the front page and # from /about, because a page nobody can find from the site is a bad page whatever a # crawler thinks -- so that argument has not stopped being true, and this line is no longer # load-bearing for discovery. # # It goes in anyway, for the reason that has changed rather than the one that has not: # /docs is the URL a directory submission points at and the one page here that is aimed at # a search term this domain can win, so it is the one page whose indexing is worth being # explicit about rather than left to a crawl finding its way down from the root. Sitemap: https://tectonicplates.app/sitemap.xml