visual-skills
AI film director skills for agents: cinematic dramaturgy (Murch, blocking, montage) + exact prompt syntax for Seedance 2.5, Kling 3.0 Turbo/Omni, Veo 3.1, Nano
What is visual-skills?
AI film director skills for agents: cinematic dramaturgy (Murch, blocking, montage) + exact prompt syntax for Seedance 2.5, Kling 3.0 Turbo/Omni, Veo 3.1, Nano Banana 2, GPT Image 2.5.
Overview of visual-skills
What it is
Visual Skills is a set of two Claude Skills that turn an AI agent into a film crew. The video skill writes AI video prompts using cinematic dramaturgy (editing, staging, montage) and exact syntax for supported video models. The image skill writes image prompts using art direction principles. Both skills select the appropriate model for the task, apply its exact syntax, and return a copy-paste-ready prompt. The skills are built around a dramaturgy-first approach: model syntax is worthless without proper dramaturgy. The core reference file is video/references/dramaturgy.md, which contains hard rules for how films are built, compressed into rules an agent can execute on a 5-30 second clip. The skill enforces seven laws: scene formula (desire + obstacle + geometry + gaze + rhythm), the Details Law (three physical facts per shot), Walter Murch's Rule of Six for editing, the three-jobs rule, blocking/camera/environment rules, montage rhythm, and the shot card with five anchors. Output is gated by a six-point dramaturgy check and a three-detail audit on every shot; a prompt that fails either is not returned.
Key features
- Two Claude Skills:
videofor AI video prompts with cinematic dramaturgy,imagefor image prompts with art direction. - Supports exact syntax for Seedance 1.0, 1.5 Pro, 2.0, 2.0 Mini, 2.5; Kling 1.6–2.6 Pro, 3.0, Turbo, Omni; Veo 3, 3.1; Nano Banana 2 Lite, 2, Pro; GPT Image 2.5 Flare, 2.5 Sunburst, legacy 2.
- Enforces seven cinematic laws: scene formula, Details Law, Murch's Rule of Six, three-jobs rule, blocking/camera/environment, montage rhythm, shot card with five anchors.
- Output is gated by a six-point dramaturgy check and a three-detail audit; prompts that fail are not returned.
- The two skills can chain:
imagebuilds character sheets and keyframes,videoturns them into motion. - Works with the sibling skill
creative-directorfor ideas and scripts.
Use cases
Writing AI video prompts that follow cinematic dramaturgy and exact model syntax. Writing AI image prompts with art direction principles. Creating storyboards, animatic keyframes, director treatments, and prompt audits. Chaining image and video skills for character continuity across a series. Generating prompts for supported video and image models.
Getting started
See the repository README for installation and usage instructions.
License and pricing
License: CC-BY-4.0. Pricing: open source.
Based on the project's public README and repository metadata.