Changelog

Benchmark Strips Platform-Wide, Leaderboard Domain Compare, and Quality Audit Fixes

Added standard 4-column benchmark strips to 6 tools, shipped a "Compare My Domain" inline widget on the Leaderboard, swept cross-tool trigger cards to 7 more tools, fixed onboarding error-path false progress, and corrected two data-integrity bugs (Math.random scatter → deterministic hash, competitor gap false specificity).

• Leaderboard — "Compare My Domain" widget — after seeing who's winning AI citations for a topic, users can now enter their domain inline to instantly see if they rank, their Share of Voice vs #1, and get direct deep-links to Citation Gap and Action Plan. CTA previously linked to a non-existent ?topic= param — now wired to live in-page comparison.
• Benchmark strips added to 6 tools — Backlinks, AI Citation Traffic, SERP Features Tracker, Content Editor, Speakable Validator, and Bing Copilot Citations now show Median / Top 25% / Leaders / Your Score 4-column strip alongside every score ring.
• Cross-tool trigger cards swept to 7 tools — passage, rank-tracker, kw-velocity, entity-registry, citation-causality, competitor-citation, and gsc-ai-overview now fire crTriggerCards() and crNextSteps() consistently with the rest of the platform.
• Onboarding error-path fix — when the initial domain scan failed, fake progress bars (animated to 30–70% during loading) were left at partial-fill state before the fallback results rendered. Bars now reset to 0% on error before proceeding.
• Citation Causality shadow scatter — deterministic jitter — chart previously used Math.random() for Y-axis jitter, causing data to change on every page reload. Replaced with a char-sum hash function so the chart is stable and reproducible.
• Competitor Citation per-query gap — false specificity removed — "Why competitor wins" collapsible was showing the same global signal list for every query regardless of per-query data. Now hidden unless signalDiff or compAdvantages actually exist in the API response for that query.

Critical Sidebar Fix, WebMCP Rebuild, and Tool Stability

Resolved a critical sidebar crash affecting all authenticated users, fully rebuilt the WebMCP tool to world-class standards, and fixed three tool APIs shipped earlier this week.

• Sidebar crash fix (critical) — _sidebar.js was calling getToken() in two places (sidebar build and AI disclosure modal) after that function was removed. Both calls replaced with getSession().access_token. The sidebar was silently broken for all logged-in users. Now resolved platform-wide.
• WebMCP Generator — full platform integration — tool was missing its sidebar entirely (no <aside class="cr-sidebar"> in DOM). Added sidebar, fixed layout to proper cr-layout two-column structure, corrected auth token reading (logged_in boolean → access_token JWT string), added Authorization: Bearer header to API calls, and wired crTriggerCards for cross-tool continuity.
• Speakable Validator fixes — removed hard "selectors array required" error; API now auto-discovers SpeakableSpecification CSS selectors from page JSON-LD and falls back to semantic element defaults. Frontend field references corrected (d.score, d.selectors, d.passages).
• Bing Copilot Citations fix — modal display bug (two conflicting display values in one style attribute) caused overlay to always be visible. Removed the duplicate.
• AI Visibility Plan fix — renderPlan() was reading week data from the wrong nesting level (d.weeks vs d.plan.weeks). Normalized field access with fallback chain; per-task checkboxes now persist to localStorage.
• SEARCH_TOOLS index — Speakable Validator, Bing Copilot Citations, and AI Visibility Plan added to the global Cmd+K search index.

3 New Tools: Speakable Validator, Bing Copilot Citations, AI Visibility Plan

Three previously API-only capabilities now have full tool pages, sidebar entries, and cross-tool integration — surfacing signals no competitor currently exposes.

• Speakable Validator — validates SpeakableSpecification CSS selectors against live HTML. Shows which selectors pass (with element count + passage preview), auto-discovers candidate selectors from page structure, and generates a ready-to-paste JSON-LD Speakable schema block. The only tool in the market that validates this signal end-to-end.
• Bing Copilot Citations — keyword-level Bing Web Search API check (top-10 organic presence per keyword). Microsoft Copilot grounds responses in Bing's index; this is the best available proxy for Copilot citation eligibility. AI synthesis via Gemini returns a Bing pattern verdict, Copilot citation logic explanation, and single priority action.
• AI Visibility Plan — personalized 30-day roadmap generated from your scan data. Reads saved CRPersist scan history and calls /api/visibility-plan to produce a week-by-week plan with score gain estimates, effort hours, and deep-links to the right fix tool. Markdown export for client delivery.
• Sidebar + tools.html — all 3 tools added to sidebar nav (Speakable → Audit & Technical, Bing Copilot → Citations, AI Visibility Plan → Research & Planning) and the full tool directory.
• Sitemap — 3 new URLs added (total: 28 URLs).

SEO Landing Pages, Score Projector, & Viral Share Upgrade

Seven semantic HTML5 SEO landing pages for top AI visibility search terms, an inline Score Projector card in the full auditor, and a viral share page upgrade with animated score ring and one-click social sharing.

• 7 SEO landing pages — semantic HTML5 pages for high-volume AI search terms: AI Citation Checker, llms.txt Validator, AI Visibility Score, Google AI Overview Checker, Why Isn't My Website in AI Answers, Schema Markup for AI Search, and AI Citation Leaderboard. Each has server-rendered content, JSON-LD (WebPage + SoftwareApplication + FAQPage + BreadcrumbList), 5–6 FAQs, internal links, and CTAs.
• Tools directory (/tools/) — index page with 7 tool cards, ItemList schema, breadcrumb navigation. Fully SEO-readable with no JavaScript dependency for content.
• Score Projector inline in Auditor — after every full scan, a "Score Projector" card appears in the Overview pane showing current score → projected score after top 5 fixes, with per-fix point estimates and issue severity. Links to /score-projector.html for full view.
• Share page viral upgrade — share.html now renders an animated SVG score ring (CSS stroke-dashoffset transition), letter grade, and domain label. LinkedIn, X (Twitter), and "Copy link" sharing buttons injected on every shared report. Dynamic OG image (/api/og) with domain + score params for rich social card previews.
• Sitemap expanded — 8 new URLs added to sitemap.xml (total: 25) covering the full /tools/ directory.

AI Content Editor — Research-First Multi-Phase Rebuild

Complete rebuild of the Content Editor as a 5-phase research-first content builder. Content generation is now grounded in real SERP data from the start — Perplexity Sonar + DataForSEO PAA — replacing the single-phase textarea with an end-to-end workflow: Research → Build → Edit & Score → QA → Export.

• Phase 1 — Research — new /api/content-research runs Perplexity Sonar and DataForSEO SERP PAA in parallel. Returns: PAA questions (up to 8), featured snippet text and type, AI Overview detection, Perplexity's current answer + who it cites, top organic result snippets, related searches, and a Gemini-synthesized content outline.
• Phase 2 — Build — new /api/content-build generates content section-by-section with full research context. Intro engineered for featured snippet extraction; body sections each target a specific PAA question; FAQ section built from remaining PAA pairs. "Build All" sequences all sections automatically.
• Phase 3 — Edit & Score — existing 8-signal live scorer preserved with all signals intact. Shared textarea keeps build and edit in sync. Get AI Rewrite still available for any section.
• Phase 4 — QA — Gemini QA pass returns: PAA coverage (covered vs. missed), direct answer quality (STRONG / WEAK / MISSING), attribution gaps, citation risks, schema recommendation, readability verdict, and a single "top fix" action.
• Phase 5 — Export — plain text, Markdown, and HTML+FAQPage JSON-LD downloads. FAQPage schema auto-built from the FAQ section generated in Phase 2.
• Auth gates — both new APIs (/api/content-research, /api/content-build) require checkGate and use wrapWeb. Parallel Perplexity + DataForSEO calls are non-blocking.
• CRPersist integration — citation score saved to scan history after every Edit & Score run, enabling trigger cards across the platform.

Data Integrity + Security Sweep

Seven critical fixes across the platform: accurate citation data, full API authentication on all agent endpoints, and correct scan history persistence.

• Video Intel — Perplexity accuracy fix — brand name was included in the citation check query, artificially inflating "cited" results. Query now uses only the target keyword, making citation data accurate.
• Video Intel — Crawl cap disclosure — sites with 100+ pages now show a notice that schema detection covers only the first 100 pages, preventing false confidence on large sites.
• Video Intel — AI grade label — competitive grade tile now labeled "AI Est. / schema signals" instead of implying live SERP measurement.
• Video Citation Monitor — scope bug fixed — videoUrls was referenced inside renderResults() outside its declaration scope, causing a silent ReferenceError that prevented CRPersist and trigger cards from saving. Fixed via module-level _vcmVideoUrls.
• Video Citation Monitor — scan history — CRPersist now saves citation rate + tier after every scan, enabling cross-tool trigger cards to surface VCM data.
• Content Rewrite API — auth gate — /api/content-rewrite was missing checkGate; any unauthenticated request would burn Gemini credits. Now fully gated.
• Smart Agents — auth on all 7 endpoints — all agent APIs (agent-recovery, agent-page-optimizer, agent-leapfrog, agent-brief, agent-content, agent-history, schedule-brief) now require authentication. These are the most expensive endpoints on the platform (multi-step Gemini chains); unauthenticated access was a live billing risk.
• Smart Agents — scan history — CRPersist.save() now fires on completion of Recovery, Page Optimizer, Leapfrog, and Weekly Brief agents, enabling trigger cards to surface agent run data across tools.
• SERP Features — AI Opportunity Score — every keyword card now opens with a scored interpretation card (0-100 AI Opportunity Score, HIGH/MODERATE/LOW badge, primary action with deep-link) instead of raw feature flags.
• Content Editor — dead code removed — stale rewriteSuggestion destructure cleaned from renderScore(); rewrite now always fires via dedicated /api/content-rewrite endpoint on button click.

Video Tool Suite — World-Class Redesign

All 4 video tools upgraded to world-class UX: unified design language, benchmark bands, priority action banners, underline tab bars with badge counts, and full cross-tool integration.

• Video Intel — complete redesign: 128px score ring, industry benchmark band (Developing/Growing/Leader/Elite), priority action banner surfacing top missing signal, underline tab bar with signal count badges, grouped signal sections (Critical → High → Present), bigger quick win cards with effort pills.
• YouTube Optimizer — matched Video Intel design standard: 128px ring, benchmark band (MEDIAN 42 / TOP25 65 / ELITE 82), priority banner, underline tab bar, signal count badge, clean CSS-class mode toggle. Fixed critical tab-switching bug (duplicate id attributes).
• Video Citation Monitor — Citation Coverage score ring with tier label (Strong/Growing/Not Yet Cited), industry benchmark strip (Avg 22% / Top Performers 58%), 4-tab bar (Overview / By Platform / By Query / AI Strategy) with uncited query count badge, stat tiles with colored top borders and icons, per-video citation progress bars, empty state card.
• Video Content Calendar — priority-colored week card left borders (Critical red / High yellow / default blue), citation score pills, priority banner showing first Critical video, 3-tab results (90-Day Calendar / Niche Intelligence / Export), 2-column niche intelligence grid, Export tab with Notion + Airtable copy buttons, Publishing Cadence field.
• Cross-tool Integration — all 4 video tools share unified design language, crNextSteps cross-links, crTriggerCards urgency alerts, and crLastDomain session persistence.

Platform-Wide Intelligence Sweep + Domain Hub Upgrade

Largest single-session platform upgrade: cross-tool trigger cards on all 48 tools, domain persistence across every tool, Domain Hub AI Opportunity Panel, SOV Coverage Matrix, and Shareable Report executive verdict.

• Cross-tool Trigger Cards — 17 urgency rules (critical/high/medium) reading live scan data now active on all 48 tools. Alerts surface in-context when a scan reveals a gap affecting a related tool.
• Domain Persistence — crLastDomain session-wide domain carry across all tools via data-cr-domain attribute (45 tools) and sessionStorage. Enter a domain once, it follows you everywhere.
• Domain Hub — AI Opportunity Panel — surfaces top 3 scoring gaps below 60 as clickable action cards with Critical/High/Medium impact labels and direct Fix → deep-links into the relevant tool.
• Domain Hub — Cross-tool Next Steps — crNextSteps() nav appended to all hub results with 6 tool links pre-filled with domain.
• SOV Coverage Matrix — appears in SOV History after running analysis. Scans homepage schema types for each domain (yours + competitors) in parallel and renders a 12-type coverage grid with AI Priority badges and gap highlights.
• Shareable Reports — Executive Verdict — hub reports now open with a Strong/Growing/Developing/Critical verdict banner, above/below median counts, open issues summary, and numbered Top 3 Opportunities with specific action text per signal.
• Action Plan — AI Artifacts — AI generates real brand-specific content: inferred brand name, sameAs URLs, 5 domain-specific FAQ pairs, industry classification. No more REPLACE_ placeholders.
• Mobile Responsiveness Sweep — 900px/640px/420px media query blocks in citerank.css cover 50+ grid classes across all tools. Full mobile parity.
• Video & YouTube Intel — new tool: AI citation signals for video content, Video SERP gaps, VideoObject schema blueprint, competitor comparison, 5-tab layout.
• Content Brief Generator — upgraded with Fan-Out Queries, Data Requirements, Passage Citability Targets, and 12-item AI Citation Readiness Checklist with progress bar.
• Multi-Domain Progress Dashboard — now supports up to 4 domains, 3 view modes (Signal Trends / Side-by-Side / Overlay), domain color coding, and ?comp= URL param.
• Backlink Authority — new tool: maps referring domain diversity, domain rank, and link health to AI citation probability. Citation authority score (0-100), tier badge (DOMINANT → WEAK), Gemini AI synthesis on how your link profile affects citation frequency, top 15 referring domains ranked by authority, and 4-signal explainer on why each metric matters to AI engines.
• AI Citation Traffic Value — new tool: the only tool that answers "what is my AI citation presence actually worth in traffic?" Per-engine CTR model (Perplexity 22% / ChatGPT 9% / Gemini AIO 4%), organic traffic displacement risk bands, net traffic position (AI gains vs. organic at risk), and projected clicks at 2× citation rate. Unique metric unavailable in any competitor.

Semantic Benchmark + AI Search Visibility Rebuilt

Two flagship tools upgraded to world-class depth — canvas radar chart for competitor comparison and donut market share chart for AI citation SOV, both with animated step progress, export, and shareable URLs.

• Semantic Benchmark: canvas radar chart comparing up to 4 domains across AI Readiness, Semantic Depth, E-E-A-T, and Schema Coverage axes
• Semantic Benchmark: 4-step animated progress with pulsing dots and domain color coding in table
• Semantic Benchmark: winner badges, rank-of-N display, and leader score cards with "▲ Leader" chips
• AI Search Visibility: Canvas donut chart shows AI citation market share across top 6 domains with "Others" slice
• AI Search Visibility: 4-step animated progress, Share URL, CSV export, and autorun=1 param support
• AI Search Visibility: Fixed layout to standard cr-layout/cr-main structure
• Both tools: CSV export and shareable link with pre-filled inputs and autorun

E-E-A-T Deep Dive Rebuilt + Dynamic OG Image

E-E-A-T Deep Dive upgraded to 5 tabs with a canvas radar chart, per-platform citation impact, signal audit with filtering, and a prioritized fix roadmap. Dynamic PNG OG image endpoint added for Product Hunt and social sharing.

• E-E-A-T: canvas radar chart across all 4 dimensions (Experience, Expertise, Authority, Trust)
• 5 tabs: Overview, Dimensions, Signal Audit, Fix Roadmap, Platform Impact
• Signal Audit tab: filter by found/missing, sorted by weight with dimension tags
• Platform Impact: per-platform citation scores for Google AIO, ChatGPT, Perplexity, Claude
• Share URL, CSV export, and PDF print buttons added
• Dynamic OG image API (/api/og) generates 1200×630 PNG via @vercel/og — no more SVG fallback

World-Class Tool Rebuilds: Fan-Out Analyzer, Citation Predictor, Training Data Audit

Three core tools fully rebuilt to world-class depth — multiple visualization layers, platform-specific forecasting, and comprehensive data drilldowns.

• Fan-Out Analyzer — Full Rebuild — 8 tabs: Fan-Out Tree, All Queries, Opportunities, Citation Map, Intent Map, Coverage Matrix, Topic Cloud, and Export. New SVG Venn diagram shows cross-platform citation overlap. Coverage Matrix (seed × fan-out heatmap), Intent Distribution chart with coverage rates by query type, multi-stage animated progress, sortable columns, drill-into-query, copy, share URL, and 5 export formats (CSV, opps CSV, Markdown, JSON, copy-all).
• Citation Predictor — Full Rebuild — Expanded to 5 tabs: Score Overview, Signal Breakdown, Fix List, Platform Forecast, and AI Rewrite. New platform-specific probability scores for Google AI Overviews, ChatGPT, and Perplexity with per-platform reasoning. Score breakdown by category (Structure / Entity / Schema) with visual bars. Fix List orders gaps by recoverable impact points. AI Rewrite tab: paste-and-compare editor with re-analyze flow. Gap export as CSV.
• Training Data Audit — Full Rebuild — Expanded to 5 tabs: Overview, Training Sources, Crawler Access, Knowledge Graph, and Fix List. New Training Data Score (0–100) ring. Common Crawl capture history timeline chart showing captures-per-crawl-batch. Deep Wikipedia and Wikidata panels with entity details and sameAs schema code snippets. Quick Win Checklist showing pass/fail status for all 8 key grounding signals. Share URL with autorun.
• Fixed: Fan-Out Analyzer returned "Unexpected token A" when Vercel returned an HTML error page — client now guards JSON parse in try/catch before checking res.ok.
• Fixed: gateErrorResponse in Fan-Out API was missing CORS headers — cross-origin 403 responses now properly include Access-Control headers.

Site Crawler Deep Analysis, Auth UX, First-Run Experience

Gemini-powered deep analysis now reliably fires after every crawl. Auth redirect loop fixed across all 50 pages. New-user onboarding polished for launch.

• Site Crawler Deep Analysis — Fixed — E-E-A-T scores, Topical Authority, AI Perception, and Entity Graph now populate correctly after crawl completion. Root cause was Vercel dropping fire-and-forget dispatches; fixed with waitUntil() throughout the crawl pipeline.
• Deep Analysis Page Cap — Deep analysis now targets the 60 most-linked pages (ordered by internal link count) instead of all 500, ensuring completion within Vercel's function timeout on any real-world site.
• Auth Redirect — Return to Origin — All 50 auth-gated pages now redirect back to the page you were trying to reach after signing in, instead of dropping you on the landing page.
• Site Crawler First-Run Hero — New users now see a 3-step guide (Enter domain → We audit 35+ signals → Get Gemini scores + fix list) before their first crawl, instead of an empty panel.
• Domain Projects — Improved Empty State — Clearer value prop with feature chips (weekly tracking, citation alerts, score history) and a stronger CTA.
• Fixed: crToast and crExportCsv were undefined in site-audit.html — CSV exports crashed with ReferenceError.
• Fixed: Deep analysis tabs (Topical Authority, AI Perception, Deep Analysis) showed misleading "enable on next crawl" message — now shows accurate spinner while running or clear prompt when no crawl exists.
• Removed: Client Workspaces from sidebar nav — feature is under development and caused confusion with Domain Projects.

Client Workspaces, Scheduled Reports, Intelligence Digest, P2 Scan Intelligence

Agency-grade client reporting, world-class workspace UX, and three scan intelligence upgrades that close the gap with SEMrush-level depth.

• Client Workspaces (Agency) — White-label client portal with full branding control: agency name, tagline, logo, accent color, welcome message, hide-Citerank toggle. Enable via the new 🏢 Workspace tab in Projects. Generates a stable shareable URL — clients see your brand, not Citerank's.
• Scheduled Client Reports — Configure weekly (Monday) or monthly (1st) automated HTML report emails direct to your client. "Send Test Email" button lets you verify the email before activating. Report reflects live branding settings.
• Industry Benchmark in Client Portal — Clients now see a live benchmark widget with industry dropdown selector, percentile position bar, and avg/top-25%/top-10% markers. Switches industries on the fly.
• Weekly Intelligence Digest — Enhanced — Each domain in the Monday digest now includes a Gemini-generated single-sentence priority recommendation ("This Week's Priority") plus industry benchmark context (percentile + avg score for your vertical).
• Schema Coverage ↔ Site Audit Integration — "From Site Audit" tab in Schema Coverage pulls analysis from a completed crawl job — no re-fetch. Zero extra API calls: reads from crawl_pages table directly.
• Competitor Schema Page-Level Drill-Down — Gap rows now expand to show which specific competitor pages have each schema type, with title + "view schema" link. Sourced from already-fetched data — no extra network call.
• Dimension Trend Deltas — Above the dimension chart in Scan History, color-coded pills now show current score + pp movement (▲/▼/→) since the first scan per dimension.
• Fixed: Weekly scheduled report cron was missing from vercel.json — weekly reports never fired.
• Fixed: Workspace tab project selector was broken (matched no DOM nodes) — rebuilt with dropdown from cached project list.
• Fixed: workspace_branding table missing report_frequency and report_email columns — migration 015 applied.
• Fixed: All browser alert()/confirm()/prompt() calls replaced with crToast/crConfirm/crPrompt across auditor.html, projects.html, settings.html, and schema-coverage.html.

Semantic Benchmark, Topical Authority, GSC CTR Correlation, Rank Tracker Opportunity Score

Four new depth layers: side-by-side 18-signal competitor benchmark, topic cluster authority mapping, CTR vs AI Overview impact analysis, and per-keyword opportunity classification in Rank Tracker.

• Semantic Benchmark — New Research & Planning tool. Compare your site against up to 4 competitors across 18 AI-visibility signals: Citerank score, schema types, entity density, FAQ count, structured headings, passage quality, training data coverage, E-E-A-T grade, and more. Side-by-side table highlights your best/worst relative to the field. AI synthesis section extracts your quickest wins, biggest gaps, and competitive advantages. Deep-links into schema generator and passage tools for each gap.
• Topical Authority Analyzer — New Research & Planning tool. Crawls your site (up to 100 pages), infers your topical domain, identifies coverage clusters, and scores Topical Gap: how much of the topic you actually own vs. competitors. Returns missing subtopics, thin clusters, cluster keyword lists, and a 90-day content plan to plug the gaps. One-click content plan generation after analysis.
• GSC AI Overview — CTR Impact Analysis — New section inside the GSC AI Overview tool. For every keyword where you appear in both AI Overviews and organic results, shows AI vs. organic impression share as a dual-bar visualization with risk classification (High / Medium / Low impact on organic CTR). Surfaces where AI Overviews are cannibalizing your clicks.
• AI Rank Tracker — Opportunity Score — New "Opp" column on every keyword row. Classifies positions as Top 3 (defend), Mid (P1 but not top 3 — optimize), High (page 2 — immediate opportunity), Reach (pages 3–5), or Long (51+). Color-coded badges make priority clear at a glance.
• Fixed: Industry Benchmarks "All Industries" table Top 10 column showed "—" for every row — API was building allIndustries without including top10 values.
• Fixed: E-E-A-T Deep Dive was using cr-page-content wrapper instead of cr-layout > cr-main — caused layout offset issues on some screen sizes.
• Fixed: Topbar breadcrumb showed "Citerank" for 40+ pages — PAGE_LABELS now covers every page in the product.

Scan History, PDF Export on Shared Reports, Public API + Training Data Audit

World-class agency workflow: dedicated scan history with trend charts, PDF download on shared reports, a fully documented public REST API, and a standalone Training Data Audit tool.

• Scan History Page — Dedicated /history.html page with score trend chart, domain filtering, scan-by-scan delta tracking, share/delete from history, and list vs. domain-grouped views. First-class tool rather than a dashboard section.
• PDF Export on Shared Reports — "Download PDF" button on every shared audit report (/share/:token). Full print CSS: white background, color-accurate score rings, page-break rules, Citerank footer. Sends from browser — no server-side rendering needed.
• Public REST API — GET /api/v1/score?domain=example.com with Bearer auth. SHA-256 hashed API keys (show-once reveal), rate limits per plan (free=10/day, agency=∞), full docs at /docs/api. Generate keys in Settings → Developer API.
• Training Data Audit — Standalone tool in AI Presence. Checks Common Crawl index, Wayback Machine snapshots, Wikipedia/Wikidata grounding, and live bot access for 6 training crawlers (CCBot, GPTBot, ClaudeBot, PerplexityBot, Googlebot-Extended, Bytespider). Verdict + Recommendations.
• RAG-Level Passage Analysis — Page X-Ray mode: heading-anchored chunking, dense embedding + BM25 hybrid retrieval simulation (0.65/0.35 weighting), 6-axis passage scoring (length fit, self-contained, answer shape, entity density, evidence, structure), "Rewrite to 90+" with Gemini.
• Citerank Deploy (Citerank Score → WordPress) — Deploy-All wizard (select → review → launch), multi-site support, per-site status strip (score, coverage, health), proof loop showing citations before/after first deploy. Renamed from WP Integration; moved to top of Core section.
• API outage resolved — All 100+ endpoints were dead (web-style vs. Node-style invocation mismatch). Fixed via api/_webcompat.js wrapWeb adapter across all 104 files. Verified 200ms response times.
• Fixed: Settings page showing Free plan when on Agency — LemonSqueezy webhook was dead during outage window; plan sync restored.
• Fixed: API key generation silently failing — missing post() helper in settings.html.
• Fixed: API docs showing wrong dimension keys (citation_presence, schema_quality etc.) — corrected to real keys: schemaEntity, discoverability, agenticProtocol, contentCitability, entityPresence.

ChatGPT + Perplexity Citations, Multi-Tool Timeline, Bulk Project Actions

Major platform upgrade — 4 new AI engines, smarter project management, and a shareable score card.

• ChatGPT Web Search Engine — Passage Citations now checks ChatGPT (gpt-4o with web_search) to see if your domain gets cited. Joins Google AIO + Perplexity as a live engine tab.
• Perplexity Sonar Engine — Checks Perplexity sonar-search for your domain across all keywords. Cited URLs + snippets shown per keyword.
• Multi-Tool Timeline Overlay — Score history chart now overlays colored markers for every Agentic, Schema, Citation, and Passage tool run on the same date axis. See how tool activity correlates with score movement.
• Project Tag Filter + Sort — Filter project grid by any tag (multi-select). Sort by: Score Change, Score High/Low, Name, Last Scanned.
• Bulk Select Mode — Select multiple projects and bulk-tag or bulk-rescan in one click.
• Score History Benchmark Line — Dashed yellow "avg" line at 50 on the score chart so you always know where you stand vs median.
• Shareable Score Card — Share button in Domain Projects drawer generates a unique link to a public score card. LinkedIn-ready dark card with score ring, grade, and CTA.
• Why This Was Cited — Each cited passage card now shows the structural signals that earned the citation: stat presence, page position, length, query alignment, citation slot.
• Fix This Passage CTA — Missed keyword cards now show ranked fix instructions derived from the competitor's winning passage, with a re-test link.
• Competitor Page Selector — Signal diff now lets you pick any specific URL pair, not just homepage vs top-ranking page.
• Export Gaps to CSV — Download all competitor gap queries as CSV from the Gap Queries section.
• Monitor Competitor Weekly — One click saves a competitor pair as a Domain Project with auto-scan enabled.
• Apply Fix Deep Links — Score Simulator fix items now link directly to the relevant Citerank tool (Schema → schema builder, Entity → brand tool, Content → Passage Citations).

Prompt Library, Viral Share Cards, Slack Alerts

Track any AI prompt over time, share audit results on LinkedIn with a real preview card, and get score-drop alerts in Slack.

• Prompt Library — Define prompts your buyers actually ask AI ("best accounting software for restaurants") and track weekly whether your domain gets cited across Gemini, Perplexity, and ChatGPT. See quoted text when cited, see who got cited instead when you didn't. The feature Otterly built their whole product around — now native to Citerank.
• Dynamic OG share image — Every shared audit report now generates a 1200×630 LinkedIn/Twitter preview card showing grade ring, score, domain, and priority issues. The viral acquisition loop now actually works.
• Share page redesign — Animated score ring, stat chips (schema count, llms.txt status, crawler access, KG presence), cleaner issue layout, stronger CTA. Reports are now worth sharing.
• "Track in Prompt Library" button — One click from any Citation Tracker query row adds it to your Prompt Library for weekly monitoring.
• Slack alert integration — Connect your Slack workspace in Settings → Integrations. Get score-drop alerts and citation change notifications directly in a channel.

Score Simulator, Drip Emails, Embeddable Badge, Score Timeline

A batch of world-class UX improvements focused on activation, retention, and virality.

• Score Improvement Simulator — shows your projected score and exactly what to fix to get there. Every audit now tells you "adding FAQ schema takes you from 42 → 61."
• Activation drip email sequence — Day 3 and Day 7 emails automatically sent to every new user. Day 3: competitor audit prompt. Day 7: industry benchmarks with your score context.
• Embeddable score badge — paste <img src="/api/badge?domain=yourdomain.com"> anywhere. Two styles: rich card and flat pill. Live score, updates weekly.
• Score trend chart on dashboard — project cards now show sparklines. Full Chart.js timeline with citation rate in tooltip.
• LinkedIn share button in the report share modal, with pre-filled copy.
• 30-Day AI Visibility Plan now appears automatically after every audit — personalized 4-week roadmap from your actual gaps.

Bot Documentation, Brand Cleanup, Skeleton Loaders

Full platform rebrand from CrawlEye to Citerank, plus polish and performance improvements throughout.

• Skeleton loaders on Citations and Brand Intelligence — scanning no longer shows a blank screen.
• Upgraded empty states with preview visualizations and compelling CTAs across all tools.
• /bot.html — full documentation page for CiterankBot, mirroring Ahrefs/Semrush practice so site admins can verify the crawler.
• All User-Agent strings updated to CiterankBot/1.0 across crawl-worker, discover, agentic, journey simulator, compare, entity gap APIs.
• Fixed settings API action mismatch (get_all → get, save → set).
• Fixed pricing page fallback when LemonSqueezy not configured — now routes to hello@citerankscore.com instead of old Webfor email.
• Fixed digest email sender and subject line (now from digest@citerankscore.com).
• robots.txt and sitemap.xml added.
• 404.html with full tool directory.

Score History, Public Share Page, Monitoring Empty State

Score trends over time, weekly snapshots, and a fully redesigned share page for audits.

• Share page rewrite — animated SVG score ring, score breakdown bars, priority issues by severity, branded CTA strip.
• Weekly score snapshots — citerank_snapshots table captures weekly scores automatically, powering the trend chart.
• OG meta tags on landing page with canonical URL.
• Upgrade CTA strip in sidebar for free users.
• Settings + Upgrade Plan added to user account menu in sidebar.

12 Tools Live, Auth, Billing, MCP Server

Full platform launch with auth, billing, and 12 AI visibility tools live on Vercel.

• Page Auditor — 18+ signal analysis including schema, entity clarity, AI crawlers, llms.txt, passage extractability, CWV.
• Citation Tracker — live citation test across Gemini, Perplexity, and ChatGPT with grounding API.
• Brand Intelligence — perception analysis with controlled AI prompt battery.
• Site Crawler — crawl up to 250 pages (Agency plan), broken links, redirect chains, orphan detection.
• Monitoring & Alerts — weekly automated scans with score drop email alerts.
• Schema Generator — AI-powered JSON-LD generation for 18+ schema types.
• llms.txt Generator — one-click AI file generation from your site content.
• Agentic Readiness — WebMCP, MCP, OpenAPI, A2A, x402 detection.
• Journey Simulator — simulate AI agent completing tasks (book, quote, contact, price).
• E-E-A-T Deep Dive — expertise, authoritativeness, trustworthiness signal analysis.
• Domain Projects — saved domains with scan history and weekly tracking.
• Citerank MCP Server — expose audit data as MCP tools for Claude Code, Cursor, and Copilot.
• Supabase auth (email + Google OAuth), LemonSqueezy billing (Pro $29/mo, Agency $49/mo).
• Gemini 2.0 Flash baked in — no user API key required.

What's in progress

• GSC AI Overview integration — real impression data from Search Console alongside your audit
• DataForSEO LLM Mentions — actual citation counts from ChatGPT, Gemini, Perplexity for target keywords
• Perception Alignment Score — controlled AI prompt battery showing what AI says about your brand vs. what you claim
• Pricing fix — Pro $49/mo, Agency $149/mo with white-label and API access