livostyle-catalog-mcp
# livostyle-catalog-mcp
> **MCP server** giving any AI agent (Claude Desktop · Cursor · Cline · Cowork) live access to **2,766+ women's fashion products** from [Livostyle.com](https://livostyle.com).
> Drop-in shopping context for shopping agents, outfit recommenders, fashion chatbots.
[](https://www.npmjs.com/package/livostyle-catalog-mcp)
[](https://opensource.org/licenses/MIT)
[](https://zenodo.org/records/20155974)
[](https://modelcontextprotocol.io)
[](https://glama.ai/mcp/servers/arturayupov/livostyle-catalog-mcp)
## Install
### Claude Desktop
Add to `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows):
```json
{
"mcpServers": {
"livostyle-catalog": {
"command": "npx",
"args": ["-y", "livostyle-catalog-mcp"]
}
}
}
```
Restart Claude. The Livostyle tools appear in the tool list — your Claude can now recommend outfits and search a real fashion catalog.
### Cursor / Cline / Cowork
Same JSON config — drop into the MCP-server section of your client's settings.
## What it does
Six tools that any LLM agent can call:
| Tool | Purpose |
|---|---|
| `search_products` | Free-text search with filters (price, rating, type, tag, in-stock). Returns ranked product list. |
| `get_product` | Full product details — all variants, sizes, colors, prices, images, reviews. |
| `list_collections` | Browse 158 curated collections (Wedding Guest, Vacation, Coachella, Boho, …). |
| `get_collection` | Products in a specific collection. |
| `recommend_outfit` | Generate complete outfit suggestion matched to occasion + budget + season. |
| `catalog_stats` | Live catalog stats (counts, ratings, top categories). |
## Try the Stylist — no install needed
The same MCP-backed brand context, running in Claude.ai as a public shareable conversation:
- 🌸 **Garden wedding guest outfit ($150 budget, June, Albuquerque)** → [view chat](https://claude.ai/share/59c161a7-ec1e-46f4-b2ef-34e480cc308f)
- 🏖️ **5-day Cabo beach vacation capsule ($250 budget)** → [view chat](https://claude.ai/share/691359e1-dba8-42a3-b7c3-892835132562)
- 🎁 **Birthday gift for a 24yo NYC quiet-luxury sister ($80 budget)** → [view chat](https://claude.ai/share/426c2ccc-6582-4378-8783-f2b8f47cec3b)
- 💬 **Additional stylist demo** → [view chat](https://claude.ai/share/2daac1a5-7950-4fd6-b644-0db470d9d144)
Each link opens a full conversation showing live product recommendations, prices, and links pulled from Livostyle’s catalog. Continue any of them with your own Claude account — the project knowledge transfers.
## Examples (what an agent can do)
```
User: "Find me a floral midi dress under $50 with at least 5 reviews"
→ Agent calls search_products(query: "floral midi", max_price_usd: 50, min_rating: 4.5)
→ Returns 3 ranked picks with prices, ratings, links
User: "What should I wear to a garden wedding in summer? Budget $100."
→ Agent calls recommend_outfit(occasion: "garden wedding guest", budget_usd: 100, season: "summer")
→ Returns 3 complete outfit suggestions with reasoning
User: "Show me everything in the Coachella collection"
→ Agent calls get_collection(handle: "coachella-outfits")
→ Returns 475+ festival-ready picks
```
## Data source
Live catalog from [Livostyle.com](https://livostyle.com), mirrored weekly via GitHub Actions to:
- 🐙 **GitHub**: [arturayupov/womens-fashion-catalog-open-data](https://github.com/arturayupov/womens-fashion-catalog-open-data)
- 🤗 **Hugging Face**: [arturayupov/womens-fashion-catalog](https://huggingface.co/datasets/arturayupov/womens-fashion-catalog)
- 📦 **NPM** (this package): [livostyle-catalog-mcp](https://www.npmjs.com/package/livostyle-catalog-mcp)
The MCP server fetches `products.json` from the GitHub mirror on first use, caches in memory, refreshes every 6 hours.
## What's in the catalog
- **2,766+** active women's fashion products
- **158** curated collections
- **4.76** avg rating, 15,937 total reviews
- 99% products with 5+ reviews · 99% with 4+ images
- Price range: $18–$120 USD
- US shipping, free over $97, 7-day returns
- Categories: Dresses, Tops, Bottoms, Outerwear, Swimwear, Activewear, Two-Piece Sets, Accessories, Shoes, Jewelry, …
## Related
- 🎬 **[Livostyle Lookbook](https://github.com/arturayupov/livostyle-lookbook)** — 50 short 9:16 outfit videos generated from this catalog. MIT, $0 cost, ready for YouTube Shorts / TikTok / Reels.
- 📊 **[Fashion-Agent-Benchmark](https://github.com/arturayupov/fashion-agent-benchmark)** — open eval framework for AI shopping agents. 200 tasks × 5 evaluators using this catalog as ground truth.
## License
MIT — see [LICENSE](./LICENSE). Free for commercial use, AI training, research. Attribution appreciated.
- 🎀 **Free interactive tool — [Wedding Guest Outfit Finder](https://arturayupov.github.io/wedding-guest-outfit-finder/)** — 5-question quiz, 4 curated picks. Vanilla HTML, MIT, fork-able.
## Why this exists
We're in the **AI shopping era**. ChatGPT, Claude, Perplexity, Gemini are becoming the new search bar. Stores that publish their catalogs as MCP servers get drop-in distribution to every agentic shopping client. This is our experiment in being there first.
If you build something with it, drop us a line at info@arcada.store — we may feature your project.
## Contact
- **Email**: info@arcada.store
- **Site**: https://livostyle.com
- **Issues**: [GitHub Issues](https://github.com/arturayupov/livostyle-catalog-mcp/issues)
TDQS
Scored across 6 tools
Each tool has a distinct and well-defined purpose: catalog_stats gives overview, get_collection and list_collections handle collections, get_product gives full details, recommend_outfit generates outfits, and search_products is the main search. No overlapping or ambiguous boundaries.
Tool names follow a consistent snake_case pattern, mostly verb_noun (get_, list_, recommend_, search_). The only minor deviation is catalog_stats (noun_noun) instead of get_catalog_stats, but it remains clear and predictable.
With 6 tools, the server is well-scoped for a fashion catalog. Each tool serves a clear role, covering search, listing, details, recommendations, and statistics without redundancy or unnecessary complexity.
The tool set covers all essential operations for a read-only catalog: searching, browsing collections, getting product details, and even outfit recommendations. There are no obvious gaps for the stated domain.