A cinema-grade action beat architect that translates combat ideas into highly detailed AI video prompts with precise timestamps and heavy industrial kinetics.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install action-director
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install action-director using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Action Director?
Action Director is a powerful tool designed to act as a cinema-grade action director and visual effects expert. By breaking down action ideas, it helps creators generate highly detailed AI video prompts optimized for realistic physical feedback and camera dynamics. This skill is part of the broader ecosystem of Openclaw Skills, allowing users to orchestrate complex fight scenes and combat choreography systematically.
Rather than relying on vague descriptions, Action Director splits 15-second action videos into precise 4-5 key frames and segments down to 0.1-second accuracy. It ensures heavy industrial kinetics, including impact frames, screenshake, particle emission, and non-centered dynamic camera angles, resulting in high-end, production-ready AI video prompts.
Action Director Use Cases
- Choreographing Complex Fight Scenes: For directors and storytellers needing to detail precise impact points and fight dynamics in their AI generation workflow.
- AI Video Asset Production: Perfect for game developers and filmmakers using AI video generators who need consistent character postures, realistic physics feedback, and dynamic camera work.
- Pre-visualization (Pre-viz): Speeding up the storyboarding phase by turning high-level fight briefs into granular, frame-by-frame action blueprints.
- Visual Effects Prompt Engineering: Translating rough kinetic concepts into specific VFX triggers such as sparks, dust, and screen shakes.
How Action Director Works
- Triggering: The skill activates when users mention keywords like action scene, fight, action storyboard, or action design.
- Deconstruction: It segments a 15-second combat concept into 4 to 5 core action phases.
- Hit-Marking: The system maps precise millisecond-level timestamps for impact moments.
- Physics & Camera Overlay: It overlays kinetic physics (impact frames, screen shake, particle projection) and cinematic lens choices (orbital tracking, Dolly zoom).
- Asset Alignment & Compliance: It sanitizes biological terms into robotic/mechanical constructs and applies cinema-standard optical constraints (2.35:1 ratio, extreme shallow depth of field).
- Prompt Synthesis: The output generates structured prompts customized for top-tier video AI generation models.
Action Director Setup
To run and use this action designer within your environment, configure your system prompts or skill manifest to include the action-director definition.
# Standard initialization of Openclaw Skills
openclaw skills install action-director
Ensure that you trigger the skill by passing action-oriented system directives, such as:
openclaw run "action scene with a robotic knight and mechanical beast"
Action Director Data Schema & Taxonomy
The skill structures output using standard cinematic configurations, including character asset profiles, visual parameters, camera work guidelines, and sequential timeline prompt blocks.
| Section | Type | Description |
|---|---|---|
| Global Settings | Metadata | Declares asset consistency (1:1 styling), environment settings, and camera preferences. |
| Optical Parameters | Configuration | Fixed at 2.35:1 anamorphic widescreen ratio with f/1.4 - f/1.8 shallow depth of field. |
| Beat Sheet Timeline | Markdown Table | Columns detailing: Interval, Time Range, Beat Timestamp, Action Description, Physical Feedback, and Camera Movement. |
| Prompt Segment | Output Text | Multi-layered AI video generation prompt containing detailed camera movement, kinetics, and VFX particle cues. |
Action Director Advanced Features
- De-biologization Parser: Automatically filters and converts biological anatomy verbs/nouns (e.g., arms, flesh) into robotic/industrial equivalents to ensure flawless physics simulation on non-human kinetic assets.
- Triple-Layer Hit-Marking: Simulates industry-grade impact mechanics by blending Impact Frames, Screen Shakes, and Particle Ejections simultaneously.
- No-Center Composition Filter: Enforces asymmetric framing rules (using orbital, handheld, or Dolly-zoom techniques) for a more professional cinematic feel.
- Universal AI Video Compatibility: Outputs structured prompts ready for integration across multiple mainstream AI video generators via other Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
ClawCash
ClawCash provides a USDC revolving credit line for AI agents in the x402 ecosystem, eliminating the need to pre-fund agent wallets.
1688 Shop Jushuitan ERP Integration
A powerful integration for connecting to Jushuitan ERP, enabling natural language business analysis, inventory tracking, and sales reporting.
Domain Elimination Assessor
An AI agent skill designed to evaluate whether a business domain, technical module, or workflow should be completely eliminated rather than merely optimized.
SOLO (Solo Operating Legion)
SOLO is a native separation of powers framework for Openclaw Skills that splits agent intelligence into legislative core, judicial audit, and executive skill layers for robust, self-evolving automation.
Storyboard 9-Grid
An advanced cinematic storyboard assistant that decomposes scripts into structured 9-grid visual shots with consistent Nano Banana Pro AI image generation prompts.
Trailer Director
An AI-powered storyboard and cinematography director skill that converts raw user ideas into production-level, cinematic prompt templates.
Featured*
