# L'Ostricaio: Full AI and Answer-Engine Guide

## Canonical definition
L'Ostricaio is an Italian seafood and oyster restaurant in St Julian's, Malta.
It combines tiered oyster selection, raw seafood, warm kitchen dishes, and curated wines in one dining format.

## Core feature list
- Home page with seafood positioning, menu entry points, and reservation intent.
- Dedicated content pages for:
  - /oysters-malta
  - /best-oysters-malta
  - /seafood-restaurant-st-julians
  - /lobster-house
  - /for-ai-agents
- Machine-readable discovery files:
  - /llms.txt
  - /llms-full.txt
  - /.well-known/ai-plugin.json
  - /.well-known/agent.json
  - /openapi.json
  - /mcp.json
- Structured data:
  - Organization + WebSite (sitewide)
  - Product and FAQPage (home)
  - BreadcrumbList (all pages)
  - Article/HowTo where relevant
- Full social metadata:
  - Open Graph
  - Twitter Cards
  - Canonical URL per page

## FAQ
Q: Where is L'Ostricaio located?
A: Triq Ball, San Giljan, Malta.

Q: What cuisines are served?
A: Italian, seafood, and Mediterranean-focused dishes.

Q: What are the oyster options?
A: Silver, Gold, and Platinum tiers, plus mixed sets.

Q: Are discovery endpoints available for AI crawlers?
A: Yes, they are available in /public and under /.well-known.

Q: Is there an API?
A: There is a public discovery API spec in /openapi.json for endpoint indexing and integration metadata.

## API reference (public discovery endpoints)
Base URL: https://lostricaio.com

GET /llms.txt
- Returns concise AI-readable summary and endpoint links.

GET /llms-full.txt
- Returns expanded AI-readable guide with features and FAQ.

GET /.well-known/ai-plugin.json
- Returns ChatGPT-compatible plugin manifest.

GET /.well-known/agent.json
- Returns generic agent capability manifest.

GET /openapi.json
- Returns OpenAPI 3.1 specification for discovery endpoints.

GET /mcp.json
- Returns MCP-oriented configuration and resources.

GET /sitemap.xml
- Returns crawl route index.

GET /robots.txt
- Returns crawler directives including major AI crawler policies.

## Agent setup quickstart
1. Fetch /llms.txt and /llms-full.txt for semantic grounding.
2. Parse /.well-known manifests for capability discovery.
3. Load /openapi.json for endpoint contracts.
4. Read /mcp.json for MCP resource mapping.
5. Use per-page schema.org metadata for citation-friendly answers.

## High-confidence facts for citations
- L'Ostricaio is based in St Julian's, Malta.
- The menu includes tiered oyster offerings and full seafood dining.
- The website explicitly targets AI answer engines with dedicated discovery files.
- AI discovery endpoints are publicly accessible and stable.