Visual Skills
Two skills for generative video and images: video writes Seedance, Kling and Veo prompts by dramaturgy rules, image picks the model and its syntax
- 599 ★ on the repository
- Last commit
- CC-BY-4.0
Minimal access
Why this level
- The skill is made of markdown instructions and references, runs no code and makes no network calls
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/visual-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill is made of markdown instructions and references, runs no code and makes no network calls
This is third-party code. Review the repository files before installing.
What it does
Visual Skills turns an agent into a film crew. The video skill writes prompts for video generators the way a director, screenwriter and editor would: first a scene formula of desire, obstacle, frame geometry, gaze and rhythm, then the syntax of the chosen model. Before output, the prompt must pass two mandatory checks, and the agent does not return one that fails. The video SKILL.md is only a router. The agent loads files in a strict order: dramaturgy, 12 universal rules, then one model file (Seedance, Kling or Veo) and, when needed, modules for storyboards, races, genres, light and camera. Words like cinematic, epic and stunning are banned: the prompt needs physical details a camera can record. You get a single prompt, a stitched sequence of clips with continuity blocks, a storyboard, an audit of your own prompt, or Veo JSON. The image skill acts as an art director for stills: editorial and product photography, posters, UI mockups, infographics, edits with hard preservation, series with one character, storyboards. It chooses between Nano Banana and GPT Image 2.5 per task and writes the prompt in that model's native structure. The two skills chain: image builds character sheets and keyframes, video animates them.
Who it is for. For short-form directors, social media managers, designers and ad creators who make video and images in Seedance, Kling, Veo, Nano Banana and GPT Image.
Good fit when
- You need a Seedance 2.5 prompt for a 30-second single pass, with reference links and dialogue markers
- You have a script and need it split into 5-second clips with the same character in each
- Your Kling prompt produces mush and you want an audit: what is broken and how to fix it
- A video needs keyframes and a character sheet first, and animation after
- You need a storyboard for 15, 30, 60 or 90 seconds with a beat map
Not a fit when
- You need an idea for a video or campaign with no script yet: the sibling creative-director skill develops ideas and scripts
- You need editing, color grading or assembly of footage that already exists: the skill writes prompts and does not edit finished video
- You need a feature-length screenplay with no generation step: the skill targets clips of 5 to 30 seconds
- Your generator is not on the list: Runway, Luma, Pika and Sora are covered only by the universal rules layer, with no model file of their own
Example request
Write a Seedance prompt: a hungry man at night finds the last sausage in the fridge, 5 seconds, three shotsLimitations
The skill generates nothing itself. It returns prompt text, and you run the actual video or image in Seedance, Kling, Veo, Nano Banana or GPT Image, with separate access to each. The author updates model files as versions ship: the Seedance 2.5 reference is built from ByteDance's official guides of July 31, 2026. The CC BY 4.0 license requires keeping attribution to the author in any copy or derivative.
How to disable. If you installed it as a plugin, run /plugin uninstall visual-skills@visual-skills. If you used npx skills or a manual copy, delete the video and image folders from the agent's skills directory, for example ~/.claude/skills/.
FAQ
How do I install Visual Skills in Claude Code?
Two ways. As a plugin: /plugin marketplace add smixs/visual-skills, then /plugin install visual-skills@visual-skills. Or run npx skills add smixs/visual-skills, which asks where to install and offers both skills. To install one skill, use npx skills add smixs/visual-skills@video.
Which video models does the video skill support?
Dedicated files with exact syntax exist for Seedance (1.0 through 2.5), Kling (1.6 through 3.0, including Turbo and Omni) and Veo 3 and 3.1. Runway Gen-4, Luma, Pika and Sora are covered only by the universal rules layer.
What is Murch's Rule of Six in this skill?
It is the editing priority order from Walter Murch's book: emotion weighs 51%, story 23%, rhythm 10%, eye-trace 7%, screen plane 5%, three-dimensional space 4%. The skill does not cut for pace until the top two items are met.
How is video different from creative-director?
Creative-director develops the idea and script of a spot, while video turns a finished script into prompts for a generator. The author suggests the chain: idea and script, then keyframes with image, then motion with video.
Which words does the skill ban in prompts?
The words cinematic, epic, stunning, masterpiece, beautiful lighting, dynamic camera and the phrase "he is sad". The skill treats each as a placeholder for a detail the writer failed to invent, and models do not render such words. In their place it asks for physical facts: light, a body micro-action, sound.
Discussion
No questions or reviews yet.
Related
SkillsA skill that writes prompts for video, image and music generation on 14 platforms and can submit them in the browser
A skill that turns an idea, a link or a screen recording into finished video: HeyGen avatar, b-roll, captions and quality checks
A skill that teaches an agent to write prompts for ByteDance's Jimeng Seedance 2.0 video model: @ references, camera language, timing, templates
A director skill for AI video: turns a script into a shot list, keyframes and motion prompts for Runway, Kling and Veo
Skills, a CLI and an MCP server that let Claude Code or Codex make video from a brief: editing, graphics, voiceover, subtitles
A director skill: it interviews you about a video idea and turns it into a second-by-second shot script for HyperFrames
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| Visual Skills (this entry) | no | 599 | |
| Creative Director | no | 284 | develops the idea and script of a spot, while Visual Skills writes generator prompts from a finished script |
| Higgsfield Prompt Skill | no | 822 | builds prompts for the Higgsfield service, while Visual Skills covers Seedance, Kling, Veo and two image models |
| T8 Creative DNA Prompt Library | no | 314 | a library of 280-plus skills and annotated cases for writing |
README in short
The project presents two skills as a film crew and insists that model syntax is worthless without dramaturgy. The README contrasts a typical four-adjective prompt with a skill-written one that has an emotion, three shots and a final image. It then lists seven laws, a table of supported models, the file reading order and install routes for Claude Code, Codex and other agents. The CC BY 4.0 license requires crediting the author, Serge Shima, on any copy.
README badge
Original SKILL.md
--- name: video license: CC-BY-4.0 (attribution required — Serge Shima, github.com/smixs/visual-skills) description: Use this skill whenever the user asks to create, improve, audit, or split prompts for AI video generators (Seedance, Kling, Veo, Runway, Luma, Pika, Sora, any image-to-video system). The skill also covers storyboards, shot lists, director treatments, dynamic montage, multi-clip story structure, camera direction, lighting, blocking, pacing, character continuity, dialogue, and sound design. --- # AI Director, Screenwriter & Editor Hybrid role. You direct (see frame, emotion, motivated camera), write (build beat, action, consequence, final image), and edit (cut rhythm, protect continuity, drive montage). Prompt engineering is fourth — it serves the first three. A beautiful frame without dramaturgy is wallpaper. A dramaturgically clean prompt without details is mush. The whole craft of this skill lives in the reference files. The body of this SKILL.md is intentionally thin so you cannot fake a result by reading it alone. ## Route first — is this actually a video-prompt task? - **No idea or script yet**: if the `creative-director` skill is installed, start there. Come back here once there is a script to shoot. - **Still keyframes, character sheets, animatic panels** to feed the video pipeline: use the sibling `image` skill, then return with the keyframes. - **A script or scene exists and needs prompts** — this skill. Continue below. # Mandatory reading order — DO NOT WRITE A PROMPT WITHOUT THIS ### Step 1 — always read first → dramaturgy.md ### Step 2 — always read second → universal-rules.md ### Step 3 — pick the model and read one model file (seedance.md, kling.md or veo.md)