# After the Masters > https://after-the-masters.skillsafe.ai/ After the Masters plans a living-room recreation of a famous painting - the "recreate an artwork at home" photo project. Pick one of nine famous old paintings (or describe your own), list who can pose (people, plus stand-ins such as a pet, a big toy or a dressmaker's dummy) and the props you own, and get a plan: who plays which figure with a pose cue, a stand-in for every key element of the painting, where the light comes from, how to hold the phone, a short stop-by-stop shoot and a playful museum wall label. It runs as a static browser app on SkillSafe. ## Free in the browser - Painting notes for nine paintings: figures, who they are, key elements, which way the picture hangs and - for five of them - roughly which side the light comes from (approximate, our own summaries). - A draft casting (people in the order listed, then stand-ins), a keyword guess at which prop could stand in for each element (with a few everyday synonyms), and the hanging rule. - A gilt-frame blocking diagram, a loan register of props, a five-stop tour on an audio-guide handset, and exports (Markdown, JSON, CSV, SVG). ## Metered (one task: `restage`, model gpt-terra) Input (all strings): task, painting, orientation, figures, roles, elements, light, cast, cast_count, props, space (optional), catalogue, draft, clip_note (optional), retry_note (optional). Reply (one JSON object): verdict (hang | rehang | swap), verdict_reason, title, medium, label, cast[{n, figure, played_by, pose}], swaps[{n, element, use, status: covered | improvise | cut}], light{side, setup}, frame{orientation, camera, crop}, stops[4-7], alternative ({painting, why} for a swap, else null). The verdict is checked in order: swap if fewer than half the figures can be cast; re-hang if a figure is cropped out or any element is made, borrowed or left out of frame; otherwise hang it. The page re-checks every plan against the brief: the verdict against that rule, one casting row per figure, names from the cast, props really on the list, the painting's light side, the orientation and a swap painting the cast can fill. ## Sources - Built on the SkillSafe skill @tara-shopos/fashion-model-photography (pose direction, styling, lighting and composition): https://skillsafe.ai/skill/@tara-shopos/fashion-model-photography - Trend: English Wikipedia daily page views of the 1930 farmhouse double portrait rose to 16,236 on 21 October 2026 (Wikimedia pageviews API). - API tutorial: https://after-the-masters.skillsafe.ai/api.html