# ContextKit > ContextKit turns messy web pages into a clean, AI‑ready bundle—language‑scoped `llms.txt`, Markdown copies of key pages and a ready‑made robots.txt snippet—so chatbots quote **your** content accurately. ContextKit is a Finnish‑born SaaS that automates creation and upkeep of LLM context files. We operate globally and have customers in almost every country in the world. Three tiers (Basic, Premium, Continuous) cover brochure sites up to large multilingual doc portals. Bundles arrive in ≤ 48 h and require only a static‑file drop—no code changes. --- ## Overview * **Problem** AIs guess if they can’t find canonical info, showing wrong prices or policies. * **Solution** ContextKit crawls your site, ranks pages with frontier LLMs, strips HTML noise, writes concise Markdown + `llms.txt`, and emails you a ready‑to‑upload ZIP. We make sure your website and services are represented in the best possible light. * **Bundle size** Core files stay **under ≈ 8 k tokens** per language; an optional `llms-full.txt` (all docs concatenated) serves 100 k‑token models. --- ## How ContextKit Works 1. **Submit & Pay** — 30‑second form + Stripe checkout. 2. **Analyse & Curate** — LLM ranking selects the most relevant pages; content is cleaned into Markdown. We carefully select what content to use in the llms.txt, since we have to fit within a reasonable token size. This is where fully automated solutions fail - the curation requires taste and expertise. 3. **Deploy & Relax** — Drop the ZIP at site root, paste one robots line; Continuous plan refreshes monthly with a diff report. --- ## Plans & Pricing (USD) | Plan | Price | Best for | Key limits | | -------------- | ------------------------------ | ------------------------ | ------------------------------------------------------------------ | | **Basic** | **\$149** one‑time | 5–20‑page brochure sites | 1 language · ≤ 200 pages · 3‑day turnaround · email support 24 h | | **Premium** | **\$349** one‑time | SaaS / API docs | 2 languages · ≤ 600 pages · 48 h turnaround · priority support 8 h | | **Continuous** | **\$99** setup + **\$79 / mo** | Sites that change often | 2 languages · 600 pages refreshed monthly · diff report | --- ## What’s Inside the ZIP * **`/llms.txt`** — curated, Markdown index of must‑read pages and information about your company and products. * **Markdown copies** — clean versions of those top pages (no nav, no ads). * **`llms-full.txt`** (optional) — entire doc set for large‑context models. * **robots.txt snippet** — allows GPTBot & ClaudeBot, hides `llms.txt` from Google. This is important. * **Diff report** — highlights what changed since last run (Continuous). * **Email support** — 24 h Basic, 8 h Premium/Continuous. --- ## Learn ### What is llms.txt?  `/what-is-llmstxt` *A plain‑language guide to the new “map” for large‑language models.* * **Definition** `llms.txt` is a Markdown file at `/llms.txt` listing the pages, docs and policies an LLM should read first—think *robots.txt for context*, not crawling. * **Origin** Proposed by Jeremy Howard (Sept 2024) and already adopted by Cloudflare, Hugging Face, WordLift and others. * **Why invented?** 1. **HTML noise** — nav bars & cookie banners waste tokens; llms.txt points to clean text. 2. **Scattered docs** — central index across sub‑domains. 3. **AI as front door** — stops chatbots guessing or quoting third‑party blogs. * **Basic structure** ```markdown # Title > One‑line blurb ## Docs - /docs/start.md : Quick start ## Policies - /terms.md : Terms of Service ## Optional ``` * **Rules of thumb** H1 title, H2 sections, < 8 k tokens, link to Markdown. * **Difference vs robots.txt** robots controls *where* bots crawl; llms.txt highlights *what* to read. * **Adoption examples** WordLift reports better Gen‑AI snippets; Search Engine Land calls it “a proposed standard for AI crawling.” * **Quick start** Create Markdown → Draft llms.txt → Upload → Add no‑index line → Test in browser. --- ### LLM SEO: Optimising Your Site for ChatGPT & Claude  `/llm-seo-howto` *A guide to shaping your content for accurate AI citations.* | Classic SEO | **LLM SEO / GEO** | | -------------------------------- | ------------------------------ | | Gatekeeper: Google/Bing index | ChatGPT, Claude, Perplexity | | Signals: backlinks, CWV | Content chunks, citation trust | | Artefacts: `sitemap.xml`, Schema | `llms.txt`, Markdown mirrors | **Core pillars** 1. Surface clean text · 2. Publish llms.txt · 3. Chunk & label · 4. Machine‑readable meta · 5. Fact hooks. **Formats LLMs love** Markdown docs, FAQ blocks, policy snippets, code samples. **Six‑step checklist** Audit Q\&A → Create llms.txt → Convert to Markdown → Chunk & label → Embed facts → Re‑test monthly. --- ### Generate an llms.txt in 2 Minutes  `/generate-an-llms-txt-in-2-minutes` Manual vs automated methods. 1. **Manual** — create a Markdown file as per template; upload. 2. **Automated** — ContextKit analyses your site and emails a full bundle in 48 h. Best practices: be concise, prioritise critical pages, keep Optional section, update with every doc change. --- ## Why Partner with ContextKit * **Strategic AI curation** — LLM ranking picks the *truly essential* pages. * **Standards‑driven** — Tracks spec updates; Continuous plan auto‑patches. * **Zero lock‑in** — Output is plain text; host anywhere. * **Secure & non‑intrusive** — Works from public URLs; never touches prod. --- ## Frequently Asked Questions * **Google rankings?** No, llms.txt targets LLMs, not classic SEO. * **Bigger than 600 pages?** Custom quotes available. * **Need server access?** No—deploy is a static file drop. * **Cancel Continuous?** Any time; you keep the latest bundle. * **Perfect AI answers?** Impossible, but curated context cuts hallucinations. --- ## Legal * **Terms of Service** — `/terms` (SLA, liabilities, Finnish law). * **Privacy Policy** — `/privacy` (GDPR compliance, retention). --- ## Contact * **Email** [support@yourotherai.com](mailto:support@yourotherai.com) * **Company** ContextKit is a product of *Your Other AI*, Helsinki, Finland. We serve customers globally. ---