Skip to main content
Glama
elasistentepepe

Ourfitmaker

README.md
# OutfitMaker MCP Server

Ask your AI assistant what to wear. OutfitMaker speaks [MCP](https://modelcontextprotocol.io), so Claude, ChatGPT, Cursor and every compatible client can get weather-aware outfit advice with one URL.

**Server URL:**

```
https://outfitmaker.ai/mcp
```

Transport: Streamable HTTP. No API key, no account, free.

## Connect in 10 seconds

**Claude (web and desktop):** Settings, Connectors, Add custom connector, paste the URL.

**Claude Code:**

```bash
claude mcp add --transport http outfitmaker https://outfitmaker.ai/mcp
```

**Cursor / Windsurf / any client that accepts remote MCP servers by URL:** same address, HTTP transport.

## Tools

| Tool | What it does |
|---|---|
| `what_to_wear_today` | Live weather for your city (today or tomorrow with `day=tomorrow`) plus a complete outfit formula for the occasion. Detects big morning-to-afternoon temperature swings and adds layering advice. |
| `suggest_outfit` | Curated outfit formulas: 7 occasions across every weather. Accepts min/max temperature and wind for layering and footwear advice. Weather words work in 5 languages. |
| `about_outfitmaker` | What connecting your real closet unlocks. |

Example, asked in Claude on a September evening:

> Tomorrow in Besancon: 13C in the morning up to 22C, clear sky.
>
> Denim jacket over a plain tee + chinos + white sneakers
>
> Heads-up: 9C swing between morning and afternoon: build on a light base with a removable layer you can shed at midday.

## The tools are generic. Your closet isn't.

These formulas are what a good stylist would tell a stranger. The real product is [OutfitMaker](https://outfitmaker.ai): photograph your wardrobe once, get daily outfit suggestions from clothes you actually own, weather-aware, previewed on a model. Free to start.

Full setup guide: [outfitmaker.ai/mcp](https://outfitmaker.ai/mcp)

Launch story: [Ask your AI what to wear](https://outfitmaker.ai/blog/ask-your-ai-what-to-wear-connect-outfitmaker-to-claude-chatgpt-and-gemini)

## Notes

- Rate limited per IP (fair use).
- No account data is exposed: the public tools are anonymous by design.
- Server implemented in Rails with the official [mcp gem](https://github.com/modelcontextprotocol/ruby-sdk).