# Schema ## Payload — one observation, four tar members The tars are **uncompressed**, members of one observation are **consecutive**, and the key is everything before the first `.` in a member name. Keys contain no dots. ``` .png screenshot, PNG, never re-encoded .leaf.json list of visible elements .screentag.txt serialized markup .record.json sanitized metadata ``` `` is `__`, e.g. `shard-0106__scene-002ad3e5b23cbfd9-step00`. It is shard-qualified because a few scenes were captured in two shards, so a stem alone is not unique; `(shard, stem)` is. ## `leaf.json` — an element ```jsonc { "type": "Button", // release class, 55 in the schema "role": "push button", // AT-SPI role it came from "name": "Save", "visible_text": "Save", // text a reader can actually see "visible_text_status": "visible", "rect": {"x": 312, "y": 88, "w": 64, "h": 28}, // amodal: the whole element "visible_fragments": [ // modal: pixels it occupies {"x": 312, "y": 88, "w": 40, "h": 28} ], "is_occluded": true, "occlusion_state": "partially_occluded", "app_name": "homebank", "reading_order_index": 41, "_window_stack_index": 2, "uid": "80b06ebf5db974e2" // stable within an episode; action targets use it } ``` **`rect` is amodal and `visible_fragments` is modal.** A box drawn from `rect` alone claims pixels another window is covering. Train grounding on the fragments; use `rect` when the question is where the whole element *is*. ## `screentag.txt` One flat string per screenshot: `×4 [state] [text]`, nested to mirror the element tree. Coordinates are on a **0–500** grid, normalized to the viewport (`record["screentag_grid"]`). ## `record.json` Scene metadata for the observation. Carries `observation_key`, `scene_id`, `episode_id`, `step_index`, `split`, `group`, `width`, `height`, `apps`, `theme`, `scene`, `n_elements`, `n_windows`, `window_stack`, `occlusion`, `flags`, `provenance`, and — on any observation that an action led to — `action_into_this_state` and `effect_of_that_action`. `group` is `ep` for a frame of a click episode and `st` for a standalone scene. ## Indexes | file | one row per | notable columns | | --- | --- | --- | | `index/scenes.parquet` | scene | `split`, `split_source`, `apps`, `theme`, `resolution` | | `index/observations/.parquet` | observation | `tar_path`, `*_member`, `occluded_ratio`, `n_elements`, `n_windows`, the flags | | `index/transitions/.parquet` | transition | `before_key`, `after_key`, action fields, `effect`, `exclusion_reasons` | | `index/instructions/.parquet` | transition with an instruction | `primary_instruction`, `instruction_variants`, `referring_expression` | | `index/episodes/.parquet` | episode | `observation_keys`, `transition_ids`, `episode_status` | | `index/shard_members.parquet` | tar member | `byte_offset`, `byte_size`, `sha256` | | `checksums/shard_stats.parquet` | tar | `bytes`, `sha256`, `observations` | `split_source` is `extended` for the 12,515 scenes that are published but not state-train-eligible, and `v3` for all others. ## Transitions An action is stored on the step it **produced**: ``` observation[k-1] -- steps[k].action --> observation[k] ``` Step 0 has no action. Reading it the other way pairs every action with the screen it was not taken on, and nothing downstream would notice. Coordinates: `action_point_px` is authoritative. `action_point_norm_1000` and `action_point_screentag_500` are derived from it and agree to within one grid cell; `action_target_bbox_px` is `[x0, y0, x1, y1]`. `effect` counts what changed between the endpoints — `changed`, `magnitude`, `appeared`, `disappeared`, `moved`, `text_changed`, `state_changed`, `newly_occluded`, `revealed`, `semantic_changes`, `persisted`. `exclusion_reasons` is empty when `transition_train_eligible` is true and otherwise names every rule the transition failed. ## Instructions `index/instructions/.parquet` has one row per transition with an instruction, joined to the corpus by `transition_id`, `before_key` and `after_key`: | column | content | | --- | --- | | `transition_id`, `split`, `before_key`, `after_key`, `target_uid` | the transition and its target element | | `action_type`, `action_point_px`, `action_target_bbox_px` | the recorded click, as in `index/transitions` | | `primary_style`, `primary_instruction` | the main instruction: `standard` where it exists, else `detailed_contextual` | | `instruction_variants` | every phrasing, as `style` and `text` | | `referring_expression` | how the target is identified on the before screen | Instructions are coordinate-free: they name a goal, never a position. ## `demo/preview.parquet` A stratified sample of observations, one row per observation with the image inline, for browsing without the shards. It is stratified over split, theme, resolution, element density, occlusion band, static against episode, changed against no-op, and common against rare application. **Its images are downscaled to 1024 px on the long edge**; `data/` is canonical. ScreenTag is on a 0–500 grid normalized to the viewport, so it still describes the resized image, and `width`/`height` remain the original pixel dimensions. ## `demo/transitions/.tar` — the `transitions_preview` subset A browsing sample of recorded clicks for the Dataset Viewer, 100 per split, one per episode, varied over target application, element role, appearance preset and resolution. The key is the `transition_id`; each sample has six members: ``` .instruction.txt primary instruction for the click .before.png the before observation's PNG, byte-identical .target.png crop of the before screen around the target, outlined .after.png the after observation's PNG, byte-identical .action.txt e.g. click (1521, 688) on table column header "Compressed" in xarchiver .transition.json ids, instruction variants, action, target, effect, scene ``` `before.png` and `after.png` hash to the `sha256` of their members in `index/shard_members.parquet`. `target.png` is the only derived image: a crop of the before screen around the target, with the target box and click point drawn on it.