# Technical Files and Machine-Readable Access Canonical: https://ai-chat-hentai.com/technical/ Title: Technical Files and Machine-Readable Access | AI Hentai Chat Description: Explore the publication’s robots permissions, full text corpus, JSON and Markdown articles, sitemap, RSS and Atom feeds, publication notes and security contacts. Language: en Published: 2026-10-07 Updated: 2026-10-07 Publisher: AI Hentai Chat This is a plain-text representation of the canonical page's visible editorial content. Navigation and interactive action labels are omitted. For adults 18+. All editorial characters are AI-generated adults. No real people are depicted. ## Public content and machine-readable access The publication serves its English content as ordinary HTML in the first response. Text, headings, comparisons and navigation remain available without JavaScript, an account or an age-confirmation wall. The site covers fictional adult anime AI conversation for readers aged 18+. The files below provide alternate ways to read the same published material. They are generated from the article inventory rather than maintained as a second set of reviews. HTML remains the canonical representation; text and JSON preserve the checked dates, visible plan facts and qualifications. Crawler permissions · /robots.txt The wildcard permits every other crawler. Each named group permits all public paths. No crawl delay or content wall is imposed. Page-level indexing instructions still apply. Canonical HTML sitemap · /sitemap.xml All 30 indexable English pages, each with its own canonical address and content date. The sitemap includes the eight utility pages: publication policies, contact information and this technical resource directory. Machine-readable content guide · /llms.txt Publication scope, adult fictional-media boundaries, evidence method, categorized materials and links to HTML, Markdown and JSON representations. Full publication text corpus · /llms-full.txt All 30 indexable pages, including the eight utility pages, in one UTF-8 text document, with headings, article text, comparison tables, plan facts and visible questions. Navigation, action labels and image-control text are omitted. Structured editorial catalog · /data.json A JSON catalog with publication identity, evidence boundaries, canonical content records, dates and representation links. It describes this publication and does not expose a live conversation API. Example Markdown article · /content/home.md A plain-text representation of the homepage. Every indexable page has its own Markdown file linked from its HTML head and the content catalog. Example page JSON · /data/pages/home.json A structured representation of the homepage. Every indexable page has a corresponding JSON record with content and its canonical HTML address. Publication and production notes · /humans.txt Actual editorial responsibility, subject names, language, supplied character-media provenance and contact paths. No invented staff biographies or company credentials. Security reporting · /.well-known/security.txt The standard security-contact location, with HTTPS canonical addresses, report contact paths, English preference and an expiry date. The identical root /security.txt file is a convenience mirror. RSS 2.0 updates · /feed.xml Twenty editorial entries: comparisons, service reviews, plans and practical guides. Items carry canonical links, descriptions, permanent identifiers and dated publication metadata. Atom 1.0 updates · /feed.atom The same editorial entries in Atom, with stable identifiers, publication responsibility and English language metadata. Browser application metadata · /manifest.webmanifest Publication name, local icon and browser presentation. This is a reading site; the manifest does not imply an installed chat application. ## Explicit crawler permissions The policy contains 28 documented crawler or product-control identities. Every group includes Allow: / and an empty Disallow: . The unrestricted wildcard covers agents that do not have a separate named group. Search discovery, AI search, user-requested retrieval and documented training agents have different purposes. Some tokens control later product use rather than a separate crawler; some user-requested actions may not consult robots rules. Their roles are stated below so an access preference is not confused with a claim about what a provider will fetch or retain. Google · Googlebot : Google Search crawling and discovery. Allow: / ; empty Disallow: . Google · Google-Extended : Gemini training and grounding usage control; no separate HTTP crawler. Allow: / ; empty Disallow: . Google · Google-CloudVertexBot : Crawls requested by site owners for Vertex AI Agents. Allow: / ; empty Disallow: . Microsoft · bingbot : Bing search indexing. Allow: / ; empty Disallow: . OpenAI · OAI-SearchBot : ChatGPT Search discovery and indexing. Allow: / ; empty Disallow: . OpenAI · ChatGPT-User : User-initiated retrieval; robots rules may not apply to these actions. Allow: / ; empty Disallow: . OpenAI · GPTBot : Content collection that may support foundation-model training. Allow: / ; empty Disallow: . OpenAI · OAI-AdsBot : Validation of pages submitted as ads; this permission implies no submission. Allow: / ; empty Disallow: . Anthropic · Claude-SearchBot : Search indexing for relevant Claude responses. Allow: / ; empty Disallow: . Anthropic · Claude-User : User-initiated content retrieval. Allow: / ; empty Disallow: . Anthropic · ClaudeBot : Content collection that may contribute to model training. Allow: / ; empty Disallow: . Perplexity · PerplexityBot : Search discovery; operator states this crawler is not for foundation-model training. Allow: / ; empty Disallow: . Perplexity · Perplexity-User : User-initiated retrieval; operator says these requests generally ignore robots rules. Allow: / ; empty Disallow: . Apple · Applebot : Apple search and contextual content retrieval. Allow: / ; empty Disallow: . Apple · Applebot-Extended : Foundation-model training usage control; this token does not crawl pages. Allow: / ; empty Disallow: . Amazon · Amazonbot : Content collection for products and services; may support AI model training. Allow: / ; empty Disallow: . Amazon · Amzn-SearchBot : Search experiences, including Alexa; not foundation-model training. Allow: / ; empty Disallow: . Amazon · Amzn-User : User-initiated live retrieval; may not follow all robots directives. Allow: / ; empty Disallow: . DuckDuckGo · DuckAssistBot : Real-time retrieval for AI-assisted answers; not AI model training. Allow: / ; empty Disallow: . Mistral · MistralAI-User : User-initiated retrieval. Allow: / ; empty Disallow: . Mistral · MistralAI-Index : Automated search indexing; not generative AI training. Allow: / ; empty Disallow: . Mistral · MistralAI-Training : Content collection for generative AI training datasets. Allow: / ; empty Disallow: . Common Crawl · CCBot : Public web collection; does not require JavaScript or cookies. Allow: / ; empty Disallow: . Diffbot · Diffbot : Structured web extraction; empty Disallow is included because this crawler ignores Allow. Allow: / ; empty Disallow: . Meta · meta-externalagent : AI crawler identity documented in Cloudflare's crawler reference. Allow: / ; empty Disallow: . Meta · meta-externalfetcher : AI assistant identity documented in Cloudflare's crawler reference. Allow: / ; empty Disallow: . Meta · FacebookBot : AI crawler identity documented in Cloudflare's crawler reference. Allow: / ; empty Disallow: . ByteDance · Bytespider : AI crawler identity documented in Cloudflare's crawler reference. Allow: / ; empty Disallow: . ## How article text and structured data agree All 30 indexable pages have one canonical URL, an English title and description, one primary heading and machine-readable publication metadata. Organization and WebSite identify this independent publication. Article and WebPage records identify the current material and connect it to the website. Review directories include the actual visible service order. FAQ markup appears only where the corresponding questions and answers appear in the article. Published prices remain dated observations with their billing period and qualifications; the data does not turn those amounts into a guaranteed regional checkout quote. No invented ratings, individual authors or private conversation tests are added. The JSON and Markdown representations retain useful article structure and point back to the canonical page. The complete text corpus keeps the public copy together for retrieval without requiring a browser rendering engine. JavaScript enhancements affect presentation and actions, while the editorial text is already in the response. ## Discovery links on every page Every HTML page links to the RSS and Atom feeds in its head. Publication notes and the machine-readable guide are also discoverable there. Each indexable page additionally names its own JSON and Markdown representations, and every footer links to this resource directory and the RSS feed. The sitemap includes all 30 canonical indexable pages. Publication policies, contact information and this technical directory also permit indexing and have their own JSON and Markdown representations. These eight utility pages remain publication information rather than editorial reviews or guides. The feeds still contain the twenty actual editorial entries and exclude publication-policy pages. Robots access is open for the HTML, assets and technical files. Access permission, indexing and a generated answer are separate stages. A crawler operator determines its own indexing and citation decisions; the presence of a text guide does not certify a search position or inclusion in a model. ## Security contacts and verification scope Security reports about this publication can follow the contact instructions . The security file names the existing publication contact and a future expiry date. It does not promise a bounty, an independently audited system, a response deadline or a verified mailbox. Account and billing support for a reviewed service belongs to that service. The local delivery checks cover file presence, HTTP status, MIME types, syntax, content parity, canonical links, structured-data references, crawler permissions and browser behavior. The publication uses local assets, fonts and scripts with a restrictive content-security policy. No deployed-host, DNS, production firewall or real search-index result is asserted by those local checks. For the editorial basis of a service comparison, read how we review . For generated-character provenance, read the AI and media policy . A technical file gives software a readable description of the publication; it does not replace the evidence behind its reviews.