Start here

4 pages · about 19 min

Reading depthGetting started

Five pages, roughly twenty minutes, ending with enough understanding to integrate with the tool without re-deriving anything. If you have used a screenshot tool before, coordinates and error handling are the two pages where the model differs from what you already know; the other three extend it.

If you read three

  1. 01Coordinates and the transformIf you read one page, read this one: how the transform maps a pixel in the returned image back to a pixel on the screen.
  2. 02Your first captureA first capture end to end, with each flag and each response field read back.
  3. 03Errors that guide youEvery exit code, grouped by the decision it implies: retry, fix the request, wait, or give up.

Everything in this section