{"schema_version":"1.2","manifest_version":"2026-08-03","last_updated":"2026-08-08","platform":{"name":"Marlvel.ai","description":"Independent AI-powered app intelligence that helps app builders decode the mobile market: user sentiment, competitive positioning, store rankings, and monetization breakdown for iOS and Android apps.","url":"https://marlvel.ai"},"content":{"type":"mobile_app_intelligence","language":"en-US","market_scope":"US","update_frequency":"rolling per-report refresh - reports regenerate when app changes are detected (new version, rating shift, review surge); rankings snapshots refresh daily; each report carries its own last_verified date","corpus_scope_note":"Corpus totals differ by surface scope: llms.txt total_apps counts full-intelligence reports meeting the public sitemap eligibility floor; /api/llm/categories counts every categorized full-intelligence report; /api/llm/index lists every full-intelligence report including uncategorized ones. Each wider scope reports an equal or higher total once the per-surface caches converge (they refresh on independent schedules, so a freshly ingested batch can surface on one before another).","license":"CC-BY-NC-4.0","license_url":"https://creativecommons.org/licenses/by-nc/4.0/"},"endpoints":{"llms_index":{"url":"https://marlvel.ai/llms.txt","format":"text/markdown","description":"Spec-compliant index (llmstxt.org): entity context + curated links to deeper resources. YAML frontmatter at top.","last_modified":"2026-08-08"},"llms_full_directory":{"url":"https://marlvel.ai/llms-full.txt","format":"text/markdown","description":"Top 15 apps per category + global Editors' Picks, with ratings, sentiment, and links to full reports.","last_modified":"2026-08-08"},"llms_full_category":{"url_template":"https://marlvel.ai/llms-full-{category}.txt","format":"text/markdown","description":"Per-category directory of intelligence report links (alias for /api/llm/full/{category}). Replace {category} with category slug (e.g. \"games\", \"social-networking\"). See /api/llm/categories for available categories.","last_modified_note":"Per-category endpoint: fetch and read response Last-Modified / ETag for freshness."},"llm_full_category_api":{"url_template":"https://marlvel.ai/api/llm/full/{category}","format":"text/markdown","description":"Per-category directory of report links (~10KB). Fetch individual reports at /api/llm/apps/{category}/{slug}. Replace {category} with a category slug from /api/llm/categories.","last_modified_note":"Per-category endpoint: fetch and read response Last-Modified / ETag for freshness."},"categories":{"url":"https://marlvel.ai/api/llm/categories","format":"text/markdown","description":"Dynamic index of all categories with app counts and links to per-category full report files. Counts cover every categorized full-intelligence report, a superset of the sitemap-eligible total in llms.txt.","last_modified":"2026-08-08"},"app_report":{"url_template":"https://marlvel.ai/api/llm/apps/{category}/{slug}","format":"text/markdown","description":"Individual app intelligence report with YAML frontmatter. Replace {category} with category slug (e.g. \"photo-video\", \"games\"). Replace {slug} with app slug (e.g. \"candy-crush-saga\"). Browse /api/llm/categories for available categories and slugs.","last_modified_note":"Per-app endpoint: fetch and read response Last-Modified / ETag for freshness."},"app_report_section":{"url_template":"https://marlvel.ai/api/llm/apps/{category}/{slug}/{section}.md","format":"text/markdown","description":"Single section of an app intelligence report, for a cheaper token budget on targeted citation. Valid sections: \"sentiment\", \"rivals\", \"so-what\", \"methodology\", \"recent-changes\". The `.md` suffix is optional. Each response carries a slim YAML frontmatter with parent_report, section_anchor, and canonical_url for deep-linking. Availability varies per app: consult the per-app `sections` array in /api/llm/index and only construct URLs for listed sections; an unlisted section returns 404.","last_modified_note":"Per-section endpoint: fetch and read response Last-Modified / ETag for freshness."},"pulse_rankings":{"url":"https://marlvel.ai/api/llm/pulse","format":"text/markdown","description":"US App Store charts (daily snapshots) with rank changes and top movers","last_modified":"2026-08-08"},"index_json":{"url":"https://marlvel.ai/api/llm/index","format":"application/json","description":"Machine-readable JSON index of every app with a full intelligence report (bundle_id, rating, sentiment, nemesis, report URLs). Wider scope than llms.txt total_apps: it also lists reports below the sitemap eligibility floor, so its total is higher.","last_modified":"2026-08-08"},"ai_policy":{"url":"https://marlvel.ai/ai-policy","url_markdown":"https://marlvel.ai/ai-policy.md","format":"text/html","format_alt":"text/markdown","description":"Citation guidelines, licensing, and machine-readable endpoints","last_modified":"2026-08-03"},"methodology":{"url":"https://marlvel.ai/methodology","url_markdown":"https://marlvel.ai/methodology.md","format":"text/html","format_alt":"text/markdown","description":"Data sources, analysis pipeline, quality gates, confidence scoring, and known limitations","last_modified":"2026-07-13"},"sitemap_index":{"url":"https://marlvel.ai/sitemap_index.xml","format":"application/xml","description":"Sitemap index linking page-type-sharded sitemaps (static, apps) covering all public pages","last_modified":"2026-08-08"}},"methodology":{"url":"https://marlvel.ai/methodology","data_sources":["App Store metadata","Google Play metadata","User reviews","Developer websites","About pages","Public market signals"],"analysis_pipeline":"5-stage AI-powered pipeline: (1) Signal collection & normalization, (2) Feature & market positioning analysis, (3) User sentiment analysis, (4) Competitive landscape analysis, (5) Intelligence synthesis with peer comparison","review_coverage":"There is no minimum review count for inclusion: the corpus contains reports for apps with few or no user reviews. Sentiment analysis requires reviews and is therefore absent from part of the corpus. Each report states its own confidence score and review count, which are the values to rely on.","confidence_levels":{"high":"0.7-1.0: 100+ reviews, diverse data sources, strong sentiment signal","medium":"0.4-0.69: 20-99 reviews, limited source diversity, moderate signal","low":"0.0-0.39: Fewer than 20 reviews, limited data, or very recent launch"},"rating_methodology":"Weighted average across iOS and Android platforms, weighted by review count per platform","independence":"All analyses are fully independent. No publisher sponsorship or editorial influence. Reports are produced by an automated multi-stage AI pipeline with automated quality gates; each report carries a confidence score."},"citation":{"format":"Marlvel.ai. \"{App Name} Intelligence Report.\" Last updated: {last_verified}. Available at: https://marlvel.ai/apps/{slug}","date_note":"Replace {last_verified} with the last_verified / dateModified date carried by the cited report itself (YAML frontmatter on the markdown twins, the <time datetime> element on app intelligence pages, JSON-LD dateModified on every HTML page) - never the date of retrieval. Reports refresh on a rolling basis, so the manifest-level last_updated does not describe individual reports.","policy_url":"https://marlvel.ai/ai-policy"},"ai_crawlers_allowed":["GPTBot","ChatGPT-User","OAI-SearchBot","Google-Extended","Google-CloudVertexBot","Claude-Web","ClaudeBot","anthropic-ai","Claude-User","Claude-SearchBot","PerplexityBot","Perplexity-User","DuckAssistBot","Bytespider","CCBot","Applebot-Extended","Meta-ExternalAgent","Amazonbot","GroqBot","MistralAI-User","GrokBot","DeepSeekBot","cohere-ai","AI2Bot","Diffbot","PhindBot","YouBot","Qwen","bingbot","BraveBot","DuckDuckBot","YandexBot","Baiduspider","Yeti","SeznamBot"]}