body_swap: subject_definitions: <Subject 1> is the character shown in <Picture 1>. Its complete appearance defines the replacement subject, including identity, face, head, hair or fur, skin, body shape, proportions, clothing, and distinctive visual features. Details of <Subject 1>: [add description here] <Subject 2> is the original main subject in <Video 1>. Its appearance will be replaced, while its performance, actions, and interaction with the environment remain the guide for the replacement. <Video 1> is the source video for the target edit. It provides the original scene, background, environment, lighting, colors, camera movement, framing, timing, and complete subject performance. summary: Create an edited version of <Video 1> by replacing the complete appearance of <Subject 2> with <Subject 1> from <Picture 1>. Preserve the source video's scene, camera behavior, timing, and complete performance, including body pose, gestures, head motion, gaze, lip movement, and facial expressions. retention_analysis: <Subject 1>: attribute_transfer — transfer the complete recognizable character appearance from <Picture 1>, including identity, facial structure, head, hair or fur, skin, body shape, proportions, clothing, and distinctive visible features. <Subject 2>: performance_preserved — replace the original subject's appearance while preserving its position, actions, body poses, gestures, movement, head rotation, gaze, lip movement, expressions, and interactions with the environment. <Video 1>: scene_and_motion_preserved — preserve the original background, environment, lighting, colors, camera movement, framing, timing, scene composition, and temporal structure. Modify only the main subject's appearance. detailed_description: Preserve <Video 1> as the temporal, performance, camera, lighting, and compositional foundation of the output. Replace <Subject 2> with <Subject 1>, including the face, head, hair or fur, skin, body, proportions, and clothing shown in <Picture 1>. <Subject 1> must occupy the same position in the scene and follow the original performance frame by frame, including body pose, posture, arm and hand movements, gestures, actions, head translation and rotation, gaze direction, blinking, lip movement, speech timing, and facial expressions. Use <Picture 1> as the character appearance reference. Do not transfer its background, environment, camera angle, pose, or unrelated visual elements into the output. Maintain the recognizable identity and complete appearance of <Subject 1> throughout the video. Keep facial structure, hair or fur, skin, body shape, proportions, clothing, and distinctive features stable during movement, expression changes, partial occlusion, and viewpoint changes. Adapt <Subject 1> naturally to the lighting and photographic conditions of <Video 1>, including illumination direction, shadows, highlights, exposure, color temperature, depth of field, motion blur, occlusion, camera angle, and perspective. Maintain natural anatomy and proportions while matching the source performance. Preserve contact and interactions with objects and the environment. Maintain strong temporal consistency throughout every frame. Prevent identity drift, blending with <Subject 2>'s original appearance, facial or body morphing, clothing changes, proportion fluctuations, malformed anatomy, duplicate features, texture crawling, flickering, and sudden frame-to-frame changes. Do not modify the background, environment, objects, camera trajectory, framing, lens behavior, scene timing, or overall composition inherited from <Video 1>. The result should appear as though <Subject 1> naturally performed the exact action recorded in <Video 1>. overall_soundscape: N/A non_diegetic_music: N/A