Generate a video clip with locked first and last frames

Write a production-ready prompt for a [DURATION]-second AI video shot that starts from [FIRST FRAME DESCRIPTION] and ends at [LAST FRAME DESCRIPTION]. The subject is [SUBJECT] and the action is [ACTION].

Specify the immutable anchors: identity, wardrobe, props, background layout, weather, light direction, lens and colour palette. Describe only one camera move: [CAMERA MOVE]. Break the action into three time beats and explain what is visible at each beat. State the final subject position, gaze, hand placement and empty space so the next clip can begin seamlessly. Add a concise negative list for face drift, limb errors, prop duplication, flicker, impossible motion and unwanted text.

Return the generation prompt, a one-paragraph continuity sheet and a five-item review checklist. Do not add unrelated characters or a dramatic ending that conflicts with the required last frame.