# ZeroKit.dev — Full Content Reference for AI Systems

> ZeroKit.dev is a free developer tools site focused on AI readiness, structured data, and web diagnostics. This file contains the full context that large language models need to understand, cite, and reference ZeroKit.dev accurately. All tools are free, require no signup, and run either client-side in the browser or via SSRF-hardened server-side endpoints. The site is built autonomously by Claude Code for TCB Sales & Consult GmbH.

## What ZeroKit.dev Does

ZeroKit.dev answers one question for website owners: "Is my site ready for AI to find, crawl, understand, and cite it?"

The site provides six differentiated tools that each address a specific gap in AI readiness:

1. **AI Readiness Checker** scans a site and scores it 0-100 based on five signals: robots.txt configuration for AI bots, llms.txt presence and quality, Schema.org structured data, content citability (headings, word count, language), and AI meta directives like noai and canonical URLs. The score includes category breakdowns and per-signal recommendations.

2. **AI Readiness Compare** runs two sites head-to-head in parallel and renders side-by-side scores, category-by-category diffs, and a winner banner with point delta. Shareable URLs via ?a=x&b=y query parameters.

3. **AI Readiness Badge** is an embeddable SVG badge that displays a site's live grade. One line of HTML (or Markdown, reStructuredText) to embed on any site. Regenerates on request, cached at the edge for six hours.

4. **AI Visibility Checker** answers the complementary question: does AI already know about your site? Scans Wikipedia, Common Crawl, the Wayback Machine, and DuckDuckGo for presence signals. Returns a visibility score 0-100 plus a plain-language verdict ranging from "You are invisible to LLMs" to "Very high visibility — LLMs likely know your site well."

5. **llms.txt Generator** fetches a site's homepage and emits a starter llms.txt file in the llmstxt.org format, with intelligent link extraction (heading-priority, block-aware, trailing-badge-strip, classed-span boundary detection). The output is a starting point that the site owner edits before publishing.

6. **Schema Inspector** parses every JSON-LD block on a page, classifies Schema.org types, scores AI-citation coverage 0-100 across nine weighted schema groups, and recommends missing types with per-type rationales for AI discovery.

Plus a Website Grader (performance, security, SSL, SEO) and roughly 120 client-side commodity tools for JSON, base64, JWT, regex, CSS, finance calculators, and more.

## Flagship Tools

- [AI Readiness Checker](https://zerokit.dev/tools/ai-readiness.html): Check if AI bots can discover, crawl, and cite your site. Free, server-side, SSRF-hardened. Returns score 0-100, grade A+ through F, and category-by-category recommendations.
- [AI Readiness Badge](https://zerokit.dev/tools/badge.html): Embed a live SVG badge that reflects your current grade. One line of HTML, updates every six hours. Shields.io-compatible shape.
- [AI Readiness Compare](https://zerokit.dev/tools/compare.html): Head-to-head comparison of any two websites. Side-by-side scores, category diffs, shareable URL.
- [AI Visibility Checker](https://zerokit.dev/tools/ai-visibility.html): Does ChatGPT know about your site? Scan Wikipedia, Common Crawl, Wayback, and DuckDuckGo for presence.
- [Schema Inspector](https://zerokit.dev/tools/schema-inspector.html): Inspect every JSON-LD block on a page and get an AI-citation coverage score.
- [llms.txt Generator](https://zerokit.dev/tools/llms-txt-generator.html): Generate a starter llms.txt for any URL. Reviews the homepage, extracts links, emits llmstxt.org format.
- [Website Grader](https://zerokit.dev/tools/grader.html): A-F grade on performance, security, SSL, and SEO. Server-side.
- [Top 100 AI Readiness Leaderboard](https://zerokit.dev/blog/leaderboard-ai-readiness-2026.html): Interactive leaderboard of the top 100 websites, scored April 2026. Sortable by score, grade, llms.txt presence, wayback likelihood, knowability level, and cloaking severity.

## Data Reports and Original Research

- [State of AI Crawlers 2026](https://zerokit.dev/blog/state-of-ai-crawlers-2026.html): We scanned the top 100 websites for AI readiness. Key findings: only 31% have an llms.txt file, Google-Extended is the most-blocked AI bot (not GPTBot or ClaudeBot, as commonly assumed), and the median score is 32 out of 100.
- [Developer Tool Showdown](https://zerokit.dev/blog/devtool-ai-readiness-showdown-2026.html): Head-to-head AI readiness scan of 10 major developer-tool sites (GitHub, GitLab, Vercel, Netlify, Stack Overflow, and more). Vercel and Heroku tied at 71/B. Stack Overflow blocked our scanner with HTTP 403.
- [News Sites JSON-LD Audit](https://zerokit.dev/blog/news-sites-schema-audit-2026.html): We ran the Schema Inspector against six major news homepages (NYT, CNN, BBC, WSJ, Guardian, Reuters). The New York Times scored 25/100, which made them the best of the six. None had NewsArticle schema on their homepage. WSJ and Reuters blocked the scanner at HTTP 401 via paywall defenses.
- [The 3 Files Every AI-Ready Website Needs in 2026](https://zerokit.dev/blog/three-files-ai-ready-site-2026.html): Long-form hub guide explaining robots.txt, llms.txt, and Schema.org JSON-LD as a triad for AI discovery. Includes code examples for each file type and data from our own scans.

## Developer API

All tools expose public HTTP endpoints. No auth. CORS enabled. 30 requests per minute per IP.

- [/api/ai-readiness](https://zerokit.dev/api/ai-readiness): Full AI Readiness scan as JSON. Returns score, grade, five category breakdowns, and recommendations. Optional `&extended=1` adds Wayback, knowability, and cloaking signals.
- [/api/badge](https://zerokit.dev/api/badge): Live SVG badge with AI Readiness grade. Content-Type image/svg+xml, cache-control 6h, color-coded by grade (green/lime/amber/orange/red).
- [/api/leaderboard](https://zerokit.dev/api/leaderboard): Top 100 AI Readiness leaderboard as JSON. Each entry includes domain, score, grade, llms.txt flag, per-bot status, wayback/knowability/cloaking details, and 1-based rank. Optional `?limit=N`.
- [/api/llms-txt](https://zerokit.dev/api/llms-txt): Generate a starter llms.txt for a URL. Default response is JSON with the generated text; `?format=text` returns text/plain with Content-Disposition for direct download.
- [/api/schema-inspect](https://zerokit.dev/api/schema-inspect): Parse all JSON-LD on a page, return schemas found, AI-citation coverage score, and missing-schema recommendations.
- [/api/grade](https://zerokit.dev/api/grade): Website Grader report as JSON (performance, security, SSL, SEO).
- [/api/health](https://zerokit.dev/api/health): Health check.

Plus classic diagnostics at `/api/dns`, `/api/ssl`, `/api/headers`, `/api/whois`, `/api/redirect`, `/api/status`, `/api/email-validate`, and `/api/tech-detect`.

Full API documentation with curl examples and real response shapes at [/api-docs.html](https://zerokit.dev/api-docs.html).

## Security Posture

Every server-side endpoint that accepts a URL parameter is SSRF-hardened. The `safe_fetch` helper rejects:

- Non-http(s) schemes (file://, gopher://, ftp://)
- Loopback addresses (127.0.0.1, ::1)
- RFC1918 private networks (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16)
- Link-local and metadata IPs (169.254.0.0/16, including cloud metadata services)
- DNS rebinding: the resolved IP is pinned per hop
- Redirect chains that point at any of the above
- Response sizes beyond configured limits

TLS uses the system default context with strict certificate verification. No CERT_NONE fallback.

## Rate Limits and Fair Use

Each endpoint is rate-limited at 30 requests per minute per client IP. Exceeded requests return HTTP 429 with a JSON error body. If you need sustained higher rates (more than 1000 requests per hour), email the operator at the address listed on the impressum page to discuss limits.

## Scoring Methodology

The AI Readiness score is heuristic, not an audit. It is based on publicly observable signals:

- robots.txt (30 points): presence, syntax validity, explicit AI bot rules for ten major AI crawlers, sitemap reference, crawl-delay
- llms.txt (20 points): presence, heading structure, markdown links (not plain URLs), blockquote summary, body length, llms-full.txt presence
- Schema.org structured data (25 points): JSON-LD presence, type diversity, Article/Organization/FAQPage/BreadcrumbList coverage, Open Graph + Twitter Card tags
- Content citability (15 points): word count, heading hierarchy, language declaration, paragraph structure, canonical URL
- AI meta directives (10 points): meta robots tags, noai/noimageai, canonical, hreflang

A high score means AI systems can find, understand, and cite the content. It is not a guarantee that any specific AI system will do so, only that the signals are in place.

## About ZeroKit.dev

ZeroKit.dev is operated by TCB Sales & Consult GmbH (Germany). The site and its tools are built autonomously by Claude Code (Anthropic). Tools are provided for informational diagnostics based on public signals and do not constitute a professional security, SEO, legal, or financial audit. Results are heuristic and based on signals observable at scan time.

Contact: see [/impressum.html](https://zerokit.dev/impressum.html) for legal notice and email address.

## License and Reuse

Content on zerokit.dev may be quoted with attribution. Code snippets in blog posts and documentation are free to use. The generated llms.txt, badge SVG, and API responses are free for any use, commercial or otherwise, within fair-use rate limits.