How to use the Seedance Spanish Vocabulary agent skill (seedance-vocab-es)

What the skill makes the agent do

An agent skill is a folder of instructions that an AI assistant loads when a task matches its description. This one, named seedance-vocab-es, is triggered when you ask for Spanish Seedance 2.0 prompt wording, Spanish cinematic vocabulary, or a translation of camera, lighting, action, VFX, audio or production terms into Spanish.

The skill gives the agent a table of Spanish terms (for example travelling de acercamiento, plano medio, contraluz and sonido ambiente), a compact prompt template, and rules for avoiding vague words such as épico or de alta calidad. Instead of those words, the agent is told to describe the physical elements that produce the look: camera movement, light source, material and sound.

• Translates production meaning into concise Spanish prompt lines rather than word-for-word English.
• Keeps reference file tags exactly as written, including their spelling and punctuation.
• Keeps dialogue in its original language unless you ask for a translation.

Install the skill from the ZIP

Download the ZIP for this skill from this page. The skill ships as a single file, SKILL.md. There are two ways to install it.

Option 1: unzip it into your personal skills folder, ~/.claude/skills/, or into a project's .claude/skills/ folder if you want it only for one project. Option 2: in the Claude apps, open Settings, go to Capabilities, then Skills, and upload the ZIP there.

• Personal install: ~/.claude/skills/seedance-vocab-es/
• Project install: .claude/skills/seedance-vocab-es/ inside your project folder
• Claude apps: Settings → Capabilities → Skills → upload the ZIP

Trigger the skill with natural requests

You do not need a special command. Describe what you want in ordinary language that mentions Spanish and Seedance, or asks for Spanish cinematic terms. The skill's description tells the agent when to load it, so phrasing like the examples below is enough.

Give the agent your scene decisions up front: subject, what happens on screen, the camera move, the light and the sound. The skill is written to use decisions you have already made rather than interview you again.

• "Escribe un prompt en español para Seedance 2.0: una taza de café en una mesa de madera, luz suave de ventana, cámara en travelling de acercamiento."
• "Traduce al español estos términos de cámara y luz para mi prompt de Seedance."
• "Dame vocabulario cinematográfico en español para una escena nocturna con contraluz."

Use the compact prompt pattern

The skill includes a short template that the agent fills in. It names the reference, says what must stay unchanged, gives one camera move and states the sound. Keeping prompts to one camera move and one sound cue is the main habit the skill encourages.

If you already have a reference image or other asset, its tag (for example @Image1) must be written exactly as your Seedance setup expects. The skill does not attach files; it only places the tag in the text.

Check the vocabulary and remove filler words

When a prompt leans on words like cinematográfico, épico, impresionante, mágico or de alta calidad, the skill directs the agent to replace them with the physical details behind the effect. A prompt that says the scene is épico becomes a description of a slow dolly, a low warm light source and the sound of wind, for example.

Review the Spanish yourself. The skill states that its independent Spanish review is still pending, so the wording should be treated as a working draft and checked by a native or locale-reviewed speaker if accuracy matters for publication.

• Replace vague adjectives with a camera move, a light source, a material or a sound.
• Check that reference tags are unchanged after the agent edits the text.
• Confirm that any quoted dialogue is still in its original language.

Run the prompt in your video tool

This skill only shapes prompt text. It does not connect to the Seedance video service and does not generate or render video. You need a Seedance 2.0 setup or another video tool to run the prompt. The skill's files do not include that tool.

Outcomes depend on the video model. The skill itself makes no promises about quality, consistency or credit savings, so judge the result by running the prompt and comparing it with what you asked for.

Safety notes: scripts and network access

The skill's files are a single SKILL.md with no scripts. It does not run code on your computer and does not use the network. Its effect is limited to the text the agent writes for you.

Because it is only instructions, it is low-risk to install, but you should still read SKILL.md if you want to see exactly what the agent is told to do. Check the licence before reuse: the skill is released under the MIT licence.

• Runs scripts: no
• Uses network: no
• Licence: MIT