# PlatPhorm Insights - Complete Public Reference Service: platphorm-insights Base URL: https://insights.platphormnews.com Updated: 2026-08-18T08:56:04.813Z ## Purpose PlatPhorm Insights is the public web-quality intelligence layer for the PlatPhormNews web mesh. It analyzes public websites, ranks real analysis results, validates discovery surfaces, compares URLs, tracks timeline history, and exposes API/MCP intelligence for humans and agents. Insights is not Trace, MCP Registry, Docs, Evals, BrowserOps, Sandbox, Atlas, Monitor, Sheets, Decks, or Base. Those are integrations. ## Public And Protected Boundary Public-safe access is enabled by default for: - homepage and localized public pages - analyze UI and public URL analysis - dashboard, timeline, compare, rankings, tests, FAQ, and network views - health, OpenAPI, llms, RSS, feed, sitemap, robots, manifest, and well-known files - agents.json, agent-policy.json, ai-policy.json, trust.json, and security.txt - read-only MCP introspection - bounded validation for URL, routes, sitemap, feed, OpenAPI, llms, MCP, agent policy, robots, and structured data Protected actions require PLATPHORM_API_KEY through either: - Authorization: Bearer $PLATPHORM_API_KEY - X-PlatPhorm-API-Key: $PLATPHORM_API_KEY Protected actions include bulk analysis, private history, cross-site sync, report generation to Docs/Sheets/Decks, registry mutation, trusted-domain mutation, protected model actions, and admin actions. ## Current State Storage mode: server Storage durable: true Storage status: configured Storage note: Durable AWS Aurora PostgreSQL persistence is configured in the isolated Insights schema. Model adapter status: degraded Model output policy: deterministic fallback unless a server-side provider is configured Model-assisted output available: false ## Registry And Algorithms Executable tests: 2415 Executable suites: 7 Visible methodology suites: 14 Methodology checks: 185 Test registry version: insights-quality-2026-05-10.2 Product test statuses: pass, warn, fail, skipped, degraded, unavailable, not_applicable MCP tools: 62 Ranking profiles: 12 Public routes: 77 Protected routes: 10 Discovery artifacts: 26 Agent platforms: 11 Ranking algorithm: insights-ranking-2026-05-10.2 New Internet algorithm: new-internet-score-2026-05-10.2 Agent policy: insights-agent-policy-2026-05-16.1 Ranking profiles: default, public-web, platphorm-standard, seo-aeo-geo, agent-readiness, accessibility-first, performance-first, discovery-first, public-sector-ready, ecommerce-ready, developer-tool-ready, content-site-ready Rankings are deterministic and are calculated only from completed or explicitly degraded stored analyses. Insights does not rank a URL from fake data. ## Website-Analysis Tests vs Repository Tests Website-analysis tests are product checks used for Insights reports, UI readouts, recommendations, rankings, and MCP/API outputs. They are separate from repository unit, integration, and Playwright tests. ## Product Test Statuses pass, warn, fail, skipped, degraded, unavailable, not_applicable ## Report Sections - Performance and Core Web Vitals: 270 tests - Accessibility and Inclusion: 775 tests - SEO, AEO, GEO, and Public Discovery: 370 tests - Security, Trust, and Privacy: 315 tests - Agent, API, and MCP Readiness: 265 tests - Structured Data, Feeds, and Discovery Artifacts: 10 tests - Best Practices and Route Quality: 410 tests ## Core Public UI - https://insights.platphormnews.com/en - https://insights.platphormnews.com/en/dashboard - https://insights.platphormnews.com/en/timeline - https://insights.platphormnews.com/en/compare - https://insights.platphormnews.com/en/rankings - https://insights.platphormnews.com/en/tests - https://insights.platphormnews.com/en/network - https://insights.platphormnews.com/en/faq - https://insights.platphormnews.com/en/docs ## REST API Analysis: - POST /api/v1/analyze - GET /api/v1/results/{id} - GET /api/v1/results/{id}/trace - GET /api/v1/results/{id}/recommendations - GET /api/v1/results/{id}/tests - GET /api/v1/results/{id}/ranking - GET /api/v1/results/{id}/export Comparison and timeline: - POST /api/v1/compare - GET /api/v1/compare/{id} - GET /api/v1/timeline - GET /api/v1/dashboard - GET /api/v1/stats Ranking: - GET /api/v1/rankings - GET /api/v1/rankings/network - GET /api/v1/rankings/domains - GET /api/v1/rankings/routes - GET /api/v1/rankings/{id} - GET /api/v1/ranking-profiles - GET /api/v1/ranking-profiles/{slug} - POST /api/v1/rankings/recalculate Test registry: - GET /api/v1/tests - GET /api/v1/tests/registry - GET /api/v1/tests/registry/{version} - GET /api/v1/tests/{testId} - GET /api/v1/suites - GET /api/v1/suites/{suiteId} Scoring: - POST /api/v1/scoring/readability - POST /api/v1/scoring/searchability - POST /api/v1/scoring/answerability - POST /api/v1/scoring/audience - POST /api/v1/scoring/discovery - POST /api/v1/scoring/agent-readiness - POST /api/v1/scoring/new-internet - POST /api/v1/llm-instructions Validation: - POST /api/v1/validate/url - POST /api/v1/validate/routes - POST /api/v1/validate/sitemap - POST /api/v1/validate/feed - POST /api/v1/validate/openapi - POST /api/v1/validate/llms - POST /api/v1/validate/mcp - POST /api/v1/validate/agent-policy - POST /api/v1/validate/robots - POST /api/v1/validate/structured-data - GET /api/v1/validation/{id} Agent policy: - GET /api/v1/agent-policy - GET /api/v1/agent-policy/platforms - GET /api/v1/agent-policy/platforms/{id} - GET /api/v1/agent-policy/robots - GET /api/v1/agent-policy/summary - POST /api/v1/agent-policy/evaluate Network: - GET /api/v1/network/sites - GET /api/v1/network/graph - GET /api/v1/network/compliance - GET /api/v1/network/rankings - GET /api/v1/network/reports/latest - POST /api/v1/network/analyze Reports: - POST /api/v1/reports/docs - POST /api/v1/reports/sheets - POST /api/v1/reports/decks - GET /api/v1/reports/{id} Discovery and platform: - GET /api/health - GET /api/v1/health - GET /api/docs - GET /openapi.yaml - GET /openapi.json - GET /llms.txt - GET /llms-full.txt - GET /llms-index.json - GET /humans.txt - GET /attribution.json - GET /json-ld.json - GET /rss.xml - GET /feed.xml - GET /sitemap.xml - GET /sitemap-main.xml - GET /sitemap-index.xml - GET /sitemap-network.xml - GET /sitemap-full.xml - GET /robots.txt - GET /manifest.webmanifest - GET /.well-known/mcp.json - GET /.well-known/agents.json - GET /.well-known/agent-policy.json - GET /.well-known/ai-policy.json - GET /.well-known/humans.txt - GET /.well-known/security.txt - GET /.well-known/trust.json - POST /api/mcp - GET /api/mcp - GET /api/v1/capabilities - GET /api/v1/results/{id}/insights ## MCP Endpoint: https://insights.platphormnews.com/api/mcp Protocol: JSON-RPC 2.0 Public introspection: enabled Protected calls: bulk and report-generation tools require PLATPHORM_API_KEY Tools (62): - analyze_url: Analyze a public URL across performance, Core Web Vitals, SEO, accessibility, security, carbon, and AI-era scoring. - get_analysis_result: Retrieve a public-safe analysis result by ID. - get_core_web_vitals: Get Core Web Vitals (LCP, INP, CLS, TTFB, FCP) for a public URL. - check_accessibility: Run WCAG accessibility checks for a public URL. - analyze_seo: Analyze SEO signals and recommendations for a public URL. - analyze_aeo: Analyze answer-engine optimization signals for a public URL. - analyze_geo: Analyze generative-engine optimization and LLM discoverability signals. - check_security: Analyze public HTTP security headers. - estimate_carbon: Estimate page carbon using measured transfer size. - validate_w3c: Run local HTML, ARIA, semantic, links, and metadata validation. - compare_urls: Compare 2-5 public URLs side by side. - get_ai_recommendations: Return real recommendations from an analysis ID or fresh public URL check. - get_insights_capabilities: Return the canonical Insights capability manifest, evidence policy, focus areas, and REST/MCP interfaces. - generate_insight_brief: Generate an evidence-grounded prioritized brief from a stored public-safe analysis, including confidence, limitations, strengths, and next actions. - compare_insight_briefs: Compare 2-5 stored public-safe analyses using their evidence-grounded insight briefs and deterministic overall scores. - bulk_analyze: Protected: analyze up to 10 public URLs in one request. - get_history: Return public-safe history for a URL from durable or degraded state. - get_dashboard: Return public-safe dashboard summary state. - generate_llm_instructions: Generate LLM-ready remediation instructions from real page scoring. - get_test_registry: Return the versioned Insights quality registry. - get_test_result: Return registry metadata for a single quality test. - get_ranking: Return a deterministic ranking entry by ranking ID or source analysis ID. - list_rankings: List public-safe rankings from completed or degraded stored analyses. - rank_url: Rank a URL from recent stored public-safe analysis results. - rank_domain: Rank recent public-safe analyses for a domain. - rank_network: Rank the current public-safe Insights analysis inventory. - calculate_new_internet_score: Calculate the explainable New Internet Score for a stored analysis. - validate_sitemap: Validate sitemap XML and bounded no-dead-link samples. - validate_feed: Validate RSS or Atom feed structure and public reachability. - validate_openapi: Validate public OpenAPI structure and protected-action auth documentation. - validate_llms: Validate readable llms discovery files. - validate_mcp: Validate public JSON-RPC MCP introspection. - validate_structured_data: Validate JSON-LD structured data on a public page. - validate_route_compliance: Validate bounded public route compliance for a site. - get_site_scorecard: Return a public-safe scorecard for a domain from stored analyses. - get_network_compliance: Return public-safe network compliance summary from current Insights state. - validate_agent_policy: Validate public agent-policy JSON and JA4/public-protected boundary requirements. - validate_robots: Validate robots.txt alignment with the Insights public/protected agent policy. - get_agent_policy: Return the public Insights agent-policy document. - list_agent_platforms: List the versioned, source-labeled common agent platform registry. - get_agent_platform: Return one source-labeled agent platform policy by ID. - evaluate_agent_access: Evaluate a bounded public/protected agent access decision for a method and path. - get_robots_policy: Return the generated robots.txt policy text and validation. - get_ai_policy: Return the public AI content-use policy. - get_trust_policy: Return the public trust policy with JA4 and trace handling. - get_discovery_manifest: Return public policy, discovery, sitemap, llms, OpenAPI, and MCP route inventory. - get_public_access_summary: Return canonical public/protected counts and public-safe access policy. - get_health: Return public-safe Insights health and canonical count summary. - get_info: Return public service information, versions, counts, and discovery links. - get_route_compliance: Return public route-compliance policy and bounded validation guidance. - get_discovery_compliance: Return public discovery-compliance policy for llms, OpenAPI, RSS, sitemap, MCP, and well-known files. - create_docs_report: Protected: request a Docs report export. - create_sheet_report: Protected: request a Sheets report export. - create_deck_summary: Protected: request a Decks summary export. - update_agent_policy: Protected: update agent policy registry state. - refresh_agent_platform_registry: Protected: refresh the agent platform registry after source verification. - export_agent_audit_report: Protected: export a private agent access audit report. - create_agent_policy_report: Protected: create a Docs/Sheets/Decks agent policy report. - sync_network_graph: Protected: sync network graph and public route inventory. - recalculate_rankings: Protected: recalculate stored ranking snapshots. - check_ietf_standards: Run IETF Internet Standards (STD series) checks for a public URL. Returns automated check outcomes for URI (RFC 3986), HTTP Semantics (RFC 9110), HTTP Caching (RFC 9111), HTTP/1.1 (RFC 9112), JSON (RFC 8259), UTF-8 (RFC 3629), and IPv6/DNS (RFC 8200 / RFC 3596). Also returns registry coverage, manual review items, lab-required items, and coverage gaps. IMPORTANT: Results do NOT constitute full IETF protocol conformance. Many standards (TCP, DNS, TLS, QUIC, SMTP, routing) require packet-level or lab tests beyond this tool. See conformanceCaveat in the response for the full disclaimer. - get_ietf_registry: Return the IETF Internet Standards (STD series) registry. Includes all tracked STDs, associated RFCs, normative requirements, test definitions, and coverage gaps. This is the static checked-in registry; use standards:sync:ietf to refresh from Datatracker. Resources: - insights://capabilities - insights://dashboard - insights://timeline - insights://results/{id} - insights://tests - insights://test/{testId} - insights://suites - insights://rankings - insights://ranking/{id} - insights://new-internet-algorithm - insights://network/sites - insights://network/compliance - insights://recommendations/{id} - insights://issues/{id} - insights://brief/{id} - insights://openapi - insights://llms - insights://trust-policy - agent-policy://summary - agent-policy://platforms - agent-policy://platform/{id} - agent-policy://robots - agent-policy://ai-policy - agent-policy://trust-policy - agent-policy://public-access - agent-policy://protected-access ## New Internet Score The New Internet Score is explainable and versioned. Components: - Human Experience Score - Agent Experience Score - Discovery Integrity Score - Trust and Traceability Score - Content Utility Score - API/MCP Capability Score - Accessibility and Inclusion Score - Performance and Sustainability Score - Route Reliability Score - Network Contribution Score Unavailable measurements are labeled as degraded inputs. They are not silently converted into passing scores. ## Discovery Canonical network graph: https://platphormnews.com/api/network/graph Canonical sitemap index: https://base.platphormnews.com/sitemap-index.xml Insights validates public-safe discovery surfaces with SSRF protection, bounded response sizes, timeouts, and redirect limits. Sitemaps are generated from known public routes and do not intentionally include private or mutation endpoints. ## Trace And Privacy Insights accepts W3C traceparent/tracestate and emits safe PlatPhorm trace headers. Public traces must not include PLATPHORM_API_KEY, cookies, tokens, private URLs, raw IPs, private reports, or sensitive page data. x-vercel-ja4-digest is fingerprint-adjacent metadata. Insights may capture it for trace-linked observability, bot/automation analysis, request correlation, abuse prevention, and trust scoring. Public artifacts expose only hashed or redacted indicators and never the raw value. ## Agent And AI Policy The common agent platform registry is source-labeled and versioned. Official crawler names are included only when verified against current public docs. Unknown platform-specific crawlers are represented as policy families instead of invented user-agent strings. Public-safe AI training, search, answer, crawl, API, and MCP use is allowed with attribution and freshness links. Private, protected, admin, mutation, raw request metadata, raw JA4 digest, secret, and private report access is disallowed without PLATPHORM_API_KEY and is not exposed in public artifacts. ## Model Output Model-assisted recommendations, summaries, and reports must cite deterministic evidence where possible and must not fabricate measured data. When no provider is configured, Insights returns deterministic fallback or an explicit degraded state instead of fake AI output.