Aesthetics MCP
InfiniteInsight/aesthetics-mcp
Base de 1135+ aesthetics visuais para design com Claude
Ver no GitHubServidor MCP que fornece ao Claude conhecimento de 1135+ estilos visuais nomeados da Aesthetics Wiki, permitindo gerar tokens CSS, configurações Tailwind, briefings de design e componentes UI fundamentados em tradições estéticas reais. Inclui busca semântica, busca por cores e exploração de aesthetics relacionadas.
Instalação
npx convez-ai InfiniteInsight/aesthetics-mcp1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Gerar paletas de cores e CSS tokens baseados em aesthetics específicas (Vaporwave, Dark Academia, etc)
- Buscar aesthetics por descrição de vibe ou cores para briefings de design
- Explorar aesthetics relacionadas e descobrir inspiração visual para projetos
- Criar componentes UI e configurações Tailwind alinhados com tradições estéticas
Veredito

Ferramenta especializada e bem-executada para designers usando Claude. A base de dados pré-raspada com embeddings semânticos elimina atrito na primeira execução; as 8 ferramentas MCP cobrem casos de uso reais (busca por nome, vibe, cor, exploração). Recomendado para quem trabalha com design sistemático ou precisa fundamentar escolhas visuais em tradições estéticas mapeadas.
README
A tradução para o português deste README ainda está na fila.
aesthetics-mcp
A Model Context Protocol server that gives Claude knowledge of 1135+ named visual aesthetics from the Aesthetics Wiki. Use it to generate CSS tokens, Tailwind configs, design briefs, and UI components grounded in real aesthetic traditions.
Prerequisites
- Node.js 18+
- Claude Code (CLI, desktop app, or IDE extension)
Install
git clone https://github.com/InfiniteInsight/aesthetics-mcp.git
cd aesthetics-mcp
npm install
npm run setup
Then restart Claude Code. The MCP server and aesthetics skill are now active.
The database ships pre-scraped and pre-embedded — no scraping or model download needed on first run. The embedding model (~23MB) downloads on the first suggest_aesthetics call.
Usage
The skill activates automatically on design-related requests. You can also trigger it directly:
| Intent | Example |
|---|---|
| Named aesthetic | "make it Vaporwave" |
| Vibe description | "dark moody academic candlelight" |
| Project description | "a meditation app, calm but not clinical" |
| Random inspiration | "pick a random aesthetic for me" |
| Color-first | "find aesthetics that use black and gold" |
| Explore | "what aesthetics are related to Dark Academia?" |
MCP Tools
| Tool | Description |
|---|---|
get_aesthetic | Full details + wiki text for a named aesthetic |
search_aesthetics | FTS keyword search across all 1135+ aesthetics |
suggest_aesthetics | Semantic similarity search from a project description |
list_aesthetics | Browse all aesthetics, optionally filtered by category |
random_aesthetic | Random pick, optionally filtered by category |
search_by_color | Find aesthetics by color name (e.g. "burgundy", "earth tones") |
list_categories | All 66 distinct aesthetic categories |
find_related | Resolve an aesthetic's related/adjacent neighbors |
check_db_staleness | How recently the database was last scraped |
Re-scraping
The bundled database was scraped in June 2026. To refresh it:
npm run scrape # re-scrapes all 1135 aesthetics (~30 min)
node src/db/migrate-embeddings.js # re-computes semantic embeddings (~5 min)
Development
npm test # run all 63 tests
npm run setup # re-run after pulling changes to update skill + MCP path