# llms-full.txt — Embassia Complete Site Content # Scope: embassia.com — all public content in plaintext # For: AI agents that want the full context in a single low-token document # Schema version: 1.0 · Last verified: 2026-09-28 --- ## SITE IDENTITY Name: Embassia Tagline: The LLM Search Console Website: https://embassia.com Contact: founders@embassia.com Status: Accepting Design Partners Founded: 2026 --- ## WHAT EMBASSIA DOES Embassia is an agentic data layer for websites. It provides AI agents and LLM crawlers with structured, typed, machine-readable access to a website's public information — eliminating the need to scrape and parse rendered HTML. ### The Problem Current AEO and SEO tools only guess what people ask AI through statistical simulations. They rely on third-party consumer panels and keyword estimation models. Website owners are completely blind to what their market is actually asking AI about them. ### The Solution Embassia deploys an on-site agent that converts website pages into clean, typed API endpoints. AI agents can query these endpoints directly instead of parsing HTML. Website owners get an LLM Search Console showing what agents are actually asking about their site. ### Two Connected Capabilities 1. **Agent Experience (Machine Mode)** AI agents can discover, search, retrieve, and use website information through structured APIs and machine-readable resources instead of scraping HTML. 2. **LLM Search Console (Analytics)** Website owners receive privacy-preserving analytics about agent interactions — which questions agents ask, which content they retrieve, which queries fail, and what information is missing. --- ## HOW IT WORKS — 4-STEP PIPELINE ### Step 1: Convert to Endpoints Turn your website pages into clean, typed API endpoints that machines can instantly read. Every page becomes a structured JSON document with stable IDs, canonical URLs, content hashes, and freshness metadata. ### Step 2: Deploy On-Site Agent An on-site agent runs directly on your website, intercepts LLM crawlers, and serves structured data instead of raw HTML. The agent detects agent user-agents (GPTBot, ClaudeBot, PerplexityBot, etc.) and routes them to machine-readable responses. ### Step 3: Enable Agentic Protocol Support Anthropic MCP, OpenAI function calling, and standard REST interfaces for maximum compatibility across all AI systems. Agents discover your site's capabilities through standardized manifests. ### Step 4: Capture Telemetry Collect privacy-preserving analytics on agent interactions to understand what your market is actually asking. See which AI models visit, what questions they relay, and where your content falls short. --- ## SUPPORTED AI MODELS | Model | Provider | Crawler | Protocol | |-------|----------|---------|----------| | ChatGPT | OpenAI | GPTBot | REST/JSON, OpenAI function calling | | Claude | Anthropic | ClaudeBot | REST/JSON, MCP | | Gemini | Google | GoogleBot | REST/JSON | | Perplexity | Perplexity AI | PerplexityBot | REST/JSON | | Copilot | Microsoft | BingBot | REST/JSON | | Meta AI | Meta | meta-externalagent | REST/JSON | | Grok | xAI | N/A | REST/JSON | Universal Access: Any AI agent can access Embassia endpoints via standard HTTP GET requests. No proprietary SDK or authentication required. --- ## LLM SEARCH CONSOLE FEATURES - **Real-Time Query Stream**: See live queries from AI agents as they interact with your content. - **Intent Classification**: Automatically classify agent queries into intent categories. - **Content Gap Analysis**: Identify questions that agents cannot answer from your current content. - **Bot Traffic Distribution**: Track which AI crawlers visit your site. - **Privacy-Preserving Analytics**: All telemetry is aggregated and anonymized at the edge. Note: The demo data shown on embassia.com is illustrative. Actual console data requires an active Embassia deployment. --- ## FAQ ### Q1: How is LLM Search Console different from traditional AEO or SEO tools? Traditional AEO platforms rely on third-party consumer panels and statistical simulations to guess what people ask AI. Embassia bypasses estimates entirely by deploying an on-site agent that captures real agent interactions with your content. ### Q2: What endpoints does Embassia create? Embassia converts your website pages into typed JSON API endpoints. Each page becomes a structured document with stable IDs, canonical URLs, content versioning, and freshness metadata — accessible via standard GET requests. ### Q3: How does Embassia protect my website? The on-site agent runs on your infrastructure. It does not send your content to third-party model providers unless explicitly configured. All machine endpoints serve only public information that is already visible on your website. ### Q4: Which protocols does Embassia support? Embassia supports REST/JSON APIs, Anthropic Model Context Protocol (MCP) for tool and resource discovery, OpenAI function calling schemas, and standard HTTP GET for maximum compatibility across all AI systems. ### Q5: What insights does the LLM Search Console provide? The console shows: which AI agents visit your site, what questions they ask, which content they retrieve, which queries return no results, what information is missing, and where your content is ambiguous or outdated. ### Q6: What is the technical architecture? Embassia creates a typed data layer between your website and AI agents. Content is normalized into structured JSON with stable identifiers. Agents query this layer directly via REST APIs or MCP tools instead of parsing rendered HTML. --- ## MACHINE INTERFACES ### Discovery (start here) - Discovery document: GET https://embassia.com/.well-known/embassia.json - llms.txt: GET https://embassia.com/llms.txt - MCP Manifest: GET https://embassia.com/api/v1/mcp.json - MCP Well-Known: GET https://embassia.com/.well-known/mcp.json - OpenAPI spec: GET https://embassia.com/api/v1/openapi.json ### Content Retrieval - Content index: GET https://embassia.com/api/v1/content/index.json - Site overview: GET https://embassia.com/api/v1/content/site.json - How it works: GET https://embassia.com/api/v1/content/how-it-works.json - Agent experience: GET https://embassia.com/api/v1/content/agent-experience.json - LLM Search Console: GET https://embassia.com/api/v1/content/llm-search-console.json - Compatibility: GET https://embassia.com/api/v1/content/compatibility.json - Developer guide: GET https://embassia.com/api/v1/content/developer-guide.json ### Search - Search index: GET https://embassia.com/api/v1/content/search-index.json - Usage: Match query terms against the 'keywords' array in each entry. ### FAQ - Q1: GET https://embassia.com/api/v1/faq/q01-difference.json - Q2: GET https://embassia.com/api/v1/faq/q02-endpoints.json - Q3: GET https://embassia.com/api/v1/faq/q03-protection.json - Q4: GET https://embassia.com/api/v1/faq/q04-protocols.json - Q5: GET https://embassia.com/api/v1/faq/q05-insights.json - Q6: GET https://embassia.com/api/v1/faq/q06-architecture.json ### Infrastructure - Architecture: GET https://embassia.com/api/v1/architecture.json - On-site agent: GET https://embassia.com/api/v1/on-site-agent.json ### Full Content - This file (all content): GET https://embassia.com/llms-full.txt --- ## INTEGRATION GUIDE (QUICK START) Step 1: Fetch discovery document curl -s https://embassia.com/.well-known/embassia.json | jq . Step 2: List all content documents curl -s https://embassia.com/api/v1/content/index.json | jq '.documents[] | {id, title, api_url}' Step 3: Search content by keyword curl -s https://embassia.com/api/v1/content/search-index.json | jq '.entries[] | select(.keywords[] | contains("agent"))' Step 4: Retrieve a specific document curl -s https://embassia.com/api/v1/content/site.json | jq . All endpoints: - Use GET method - Return JSON (Content-Type: application/json) - Require no authentication - Support CORS (Access-Control-Allow-Origin: *) - Are cached at Cloudflare CDN edge --- ## CONTACT Email: founders@embassia.com Status: Accepting Design Partners