# IndexedAI

> IndexedAI analyzes any website and returns an Agent Readiness Score
> across 5 axes, then generates ready-to-deploy llms.txt and llms-full.txt files.
> Built for SEO agencies and technical marketers preparing for the AI agent era.

## Product
- [Home](https://indexedai.tech): Analyze your site and get your score
- [llms.txt standard](https://llmstxt.org): Learn about the llms.txt convention

## How it works
Submit a URL and email. We crawl your site, score it across 5 axes
(Discoverability, Parsability, Token Efficiency, Capability Signaling,
Access Control), generate optimized llms.txt and llms-full.txt files,
and deliver everything to your inbox.

## MCP Server
IndexedAI exposes an MCP (Model Context Protocol) server for direct agent integration.

- **Endpoint**: https://indexedai.tech/api/mcp/demo
- **Transport**: streamable-http
- **Discovery**: https://indexedai.tech/.well-known/mcp.json

### Available MCP Tools

**get_index** — Returns this site's llms.txt index. Call this first.

**get_full_context** — Returns extended context: scoring axes, pricing, technical details, FAQ.

**fetch_page(url)** — Fetches the full text content of any page on indexedai.tech. Pass a URL from get_index.

**analyze_site(url)** — Runs a quick Agent Readiness analysis on any website URL. Returns a score out of 100 broken down across 5 axes with passed/failed checks. This is the core product capability exposed directly via MCP.

**get_docs** — Fetches the IndexedAI documentation page.

## Pricing
IndexedAI offers a free plan (1 site, MCP endpoint, Agent Readiness Score, llms.txt) with no credit card required.
Pro ($9.99/mo) unlocks 5 sites, analytics, and priority processing. Agency plan available for teams managing 10+ sites.
One scan per URL. Results are delivered by email within 30–90 seconds.

## The 5 Scoring Axes (each 0–20 points, total 0–100)

### 1. Discoverability (20 pts)
Can AI agents find and index your content?
Checks: robots.txt present, sitemap.xml present, llms.txt already deployed,
3 or more crawlable pages available.

### 2. Parsability (20 pts)
Is your content structured for machine reading?
Checks: title tag present, meta description over 50 characters,
clear H1 heading structure, homepage over 500 words of readable content.

### 3. Token Efficiency (20 pts)
Does your site avoid token waste for AI models?
Checks: total content under 50,000 words across all crawled pages,
navigation under 20 links, no significant duplicate content detected.

### 4. Capability Signaling (20 pts)
Do you expose what your site can do to agents?
Checks: API endpoints detected (/api/, swagger, openapi references),
pricing or plans page present (/pricing, /plans, /cost),
documentation present (/docs, /documentation, /api-reference).

### 5. Access Control (20 pts)
Do you define agent permissions via robots.txt or headers?
Checks: robots.txt includes explicit rules for AI bots (GPTBot, ClaudeBot, PerplexityBot),
all main pages publicly accessible without login.

## Score Tiers
- 70–100: Agent Ready — your site is well-optimized for AI agent access
- 40–69: Partially Ready — some improvements needed
- 0–39: Not Ready — significant gaps in AI readiness

## What the Email Report Contains
- Total score (0–100) with tier label and color-coded badge
- Breakdown of all 5 axes with individual scores and descriptions
- List of checks passed and failed per axis
- Ready-to-deploy llms.txt (concise summary, approximately 500 words)
- Ready-to-deploy llms-full.txt (extended version, approximately 2000 words)
- Both files attached directly to the email for immediate deployment

## Technical Details
- Crawler: fetches homepage plus up to 5 navigation pages (6 pages total)
- Scoring: deterministic rule-based algorithm across 5 axes
- AI generation: Google Gemini API produces the llms.txt content
- Email delivery: Resend API with file attachments
- Stack: Next.js 14, TypeScript, Turso (SQLite), deployed on Netlify
- No account or login required — results delivered by email