一键导入
camera-vocabulary
Shot scale, camera move taxonomy, and cinematography rules for ai_video scene planning. Load when writing storyboard scenes in Phase 1.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Shot scale, camera move taxonomy, and cinematography rules for ai_video scene planning. Load when writing storyboard scenes in Phase 1.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
You MUST use this whenever the user asks for video editing work. Use it when video-related tasks are brought up. Editing, analyzing video, or transcribing videos
Find and download real images from the web — people, logos, brand/event stills, B-roll — to add as overlay image cards or project assets. Load when the prompt asks to source or insert images (e.g. 'add a photo of X', 'find images of the IPO', 'pull a shot of the factory').
Agent-authored workflow task: analyze transcripts across all clips, pick ONE best take per script section, discard all others. Load this when you hit montaj/select_takes in a workflow.
Write a custom JSX overlay component and add it to the project's overlay track.
Agent-authored workflow task: transcribe-and-sample one long-form horizontal source, pick N self-contained clip windows, and fan each out into its own vertical (9:16) clip project with the chosen framing. Load this when you hit montaj/find_clips in a workflow.
The shared vocabulary that domain skills use to name Montaj operations. Domain skills phrase every Montaj interaction as one of these verbs; an interface skill defines how each verb is actually performed. Load this to learn the canonical verbs before authoring or reading a domain skill.
| name | camera-vocabulary |
| description | Shot scale, camera move taxonomy, and cinematography rules for ai_video scene planning. Load when writing storyboard scenes in Phase 1. |
Camera is a first-class decision per shot, not a throwaway clause. Pick camera_move and shot_scale BEFORE writing the prose prompt — they shape the shot. The runtime prefixes the Kling prompt with both as [SHOT SCALE] and [CAMERA MOVE] tags so the generator actually executes them.
| Value | Framing | When to use |
|---|---|---|
ecu | Extreme close-up: eyes, hands, small detail | Intimate emotion, critical object, texture |
cu | Close-up: face fills frame | Dialogue, reaction, emotion |
mcu | Medium close-up: chest up | Conversation, character focus |
medium | Medium: waist up | Standard dialogue, action |
cowboy | Cowboy: mid-thigh up | Character with environment context, walking |
wide | Wide: full body + room | Establishing character in space, physical action |
very_wide | Very wide: environment dominates | Location change, scale, isolation |
aerial | Overhead / bird's eye | Geography, transition, grandeur |
Grouped semantically. Draw from multiple groups to create variety.
| Value | Motion | When to use |
|---|---|---|
push_in | Camera moves toward subject | Building tension, drawing attention |
pull_out | Camera moves away from subject | Reveal, release, ending |
dolly_forward | Camera physically advances through space | Moving into a scene, entering a room |
dolly_back | Camera physically retreats through space | Departing, farewell, distancing |
| Value | Motion | When to use |
|---|---|---|
tracking | Camera follows subject laterally | Walking, running, procession |
arc | Camera sweeps in a partial curve around subject | Transition, moderate emphasis |
orbit | Camera circles the subject fully | Dramatic emphasis, hero moment, isolation |
| Value | Motion | When to use |
|---|---|---|
crane_up | Vertical camera rise | Opening grandeur, reveal from above |
crane_down | Vertical camera descent | Arrival, landing, grounding |
jib | Crane with lateral arc component | Fluid elevation change with sweep |
| Value | Motion | When to use |
|---|---|---|
handheld_drift | Organic, slightly unstable float | Intimacy, documentary feel, unease |
snorri_cam | Camera fixed to subject, world moves | Disorientation, intoxication, dream |
| Value | Motion | When to use |
|---|---|---|
gimbal_glide | Perfectly smooth lateral/forward glide | Ethereal calm, floating through space |
| Value | Motion | When to use |
|---|---|---|
whip_pan | Fast horizontal snap | Energy burst, surprise, comedic timing |
crash_zoom | Rapid lens zoom into subject | Shock, sudden focus, impact |
rack_focus | Focus shifts between foreground/background | Reveal connection, redirect attention |
| Value | Motion | When to use |
|---|---|---|
tilt_up | Vertical rotation upward | Reveal height, awe, looking up |
tilt_down | Vertical rotation downward | Reveal below, grounding, descent |
dutch_tilt | Camera tilted off-axis | Tension, unease, stylized energy |
| Value | Motion | When to use |
|---|---|---|
locked_wide | No movement, wide frame | Stillness IS the statement — must have internal motion |
locked_close | No movement, close frame | Portrait intensity — must have internal motion |
static_macro | No movement, extreme detail | Texture, small object, time passing on surface |
| Value | Motion | When to use |
|---|---|---|
pov | Camera IS the subject's eyes | Immersion, what they see |
over_shoulder | Behind subject looking at their world | Perspective, conversation, approach |
Every shot needs motion from EITHER the subject OR the camera — never both absent.
push_in, pull_out, orbit, arc, tracking, handheld_drift, crane_up/down, whip_pan, crash_zoom, rack_focus. A static subject + static camera reads as a photograph, not a video shot.tracking, locked_wide, or static_macro is enough — let the subject carry the motion.locked_wide, locked_close, static_macro) → only when stillness IS the statement AND something else in frame is visibly moving (water rippling, smoke curling, light flickering, a figure crossing behind).At least 2 of these 4 must change between every consecutive pair of shots:
ECU/crash_zoom → wide/tracking → medium/orbit reads as deliberate cinematography.
medium/push_in → medium/push_in → medium/push_in reads as one confused camera operator.
Forbidden sameness: no two adjacent shots should share both the same shot scale AND the same camera move.
Before writing scenes, pick 4-8 camera moves that define this video's camera grammar — its visual identity. Every shot's camera_move should be drawn from this pool. This prevents both monotony (all push_in) and chaos (random moves with no through-line).
Example pool for a documentary: [push_in, pull_out, crane_up, tracking, gimbal_glide, locked_wide, tilt_up, orbit]
Example pool for high-energy: [whip_pan, crash_zoom, handheld_drift, tracking, dutch_tilt, arc, pov, crane_down]
Write fields in this order: shotScale → cameraMove → then the ## Camera / ## Subject / ## Action prompt sections. Camera choice shapes what the shot looks like — don't write the description first and try to fit a move after.
When planning scenes in Phase 1, write these structured fields on each storyboard.scenes[i]:
{
"id": "scene-1",
"prompt": "...",
"duration": 5,
"refImages": ["ref1", "ref2"],
"shotScale": "wide",
"cameraMove": "tracking"
}
The kling_generate step auto-appends [SHOT SCALE] and [CAMERA MOVE] tags to the composed prompt at generation time. Do NOT write these tags in the scene prompt text.