Measure a set of screenshots against a category of the App Screenshot Index

Reads a set of screenshots with the reader the App Screenshot Index uses on the top 100 free apps of each category, and answers with the same tactic keys: each screenshot's words, colour and tactics, and each notable tactic with whether the set has it and what share of the category, of its top ten and of the rest do. The set is one of: page_id (a page made here: its drawn iPhone screenshots), app_id (an app on the workspace: the iPhone screenshots of its listing as Pagefully holds them, which works for an app that is not on the public store), or app_store_url or app_store_id (a public App Store app). No picture or address of a picture is taken. With baseline, a second set named the same way, the answer also has changes: the tactics gained, dropped and kept from the baseline to the set. The category is the app's own unless category names one of the Index's slugs. An unknown slug is 404 and nothing is read. Where the Index does not hold the app's own category the answer has the tags, category null and category_note. A screenshot is read once and kept. When every screenshot of the request has been read the answer is 200. When some have not, the answer is 202 with a job_id: poll GET /v1/jobs/{id} until it is done, then send the same request again. Reading is metered: 60 new screenshots a day for a workspace and 10 of one set, the same on every plan. Past the day's count the answer is 429 and nothing is read. A request that starts a read is also one of the token's 20 generation starts an hour. There is no score. The numbers say what leading apps do, not what causes installs.