{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"tool_ux-sniff","slug":"ux-sniff","name":"UX Sniff","type":"product","url":"https://uxsniff.com","page_url":"https://unfragile.ai/ux-sniff","categories":["data-analysis"],"tags":[],"pricing":{"model":"freemium","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"tool_ux-sniff__cap_0","uri":"capability://data.processing.analysis.ai.powered.session.replay.with.behavioral.annotation","name":"ai-powered session replay with behavioral annotation","description":"Captures and replays user sessions with AI-driven analysis that automatically identifies friction points, drop-off moments, and rage clicks. The system ingests raw session data (mouse movements, clicks, scrolls, form interactions) and applies machine learning models to flag anomalous or problematic user behaviors without manual tagging, surfacing insights like 'user clicked submit button 5 times' or 'abandoned form after 30 seconds at email field'.","intents":["I need to see exactly what users did before abandoning my checkout flow","I want AI to automatically flag sessions with high frustration signals instead of watching hundreds of replays manually","I need to understand why specific user segments are dropping off at particular pages"],"best_for":["SaaS product managers optimizing conversion funnels","E-commerce teams diagnosing checkout abandonment","Digital agencies auditing client websites for UX debt"],"limitations":["Session replay adds ~50-100KB per session to storage; high-traffic sites may hit data retention limits on free tier","AI annotations are pattern-based and may miss context-specific friction (e.g., accessibility issues for screen readers)","Privacy-sensitive data (passwords, credit cards) requires manual masking configuration; no automatic PII detection"],"requires":["Website with JavaScript enabled","UX Sniff tracking script installed (1-2 lines of code)","GDPR/privacy policy updated to disclose session recording"],"input_types":["DOM events (clicks, scrolls, form submissions)","mouse/touch coordinates","page load metrics","custom events via API"],"output_types":["video replay with timeline","structured annotations (friction points, drop-off events)","heatmap overlay on replay"],"categories":["data-processing-analysis","user-behavior-analytics"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_1","uri":"capability://data.processing.analysis.heatmap.generation.with.interaction.density.mapping","name":"heatmap generation with interaction density mapping","description":"Generates visual heatmaps showing click, scroll, and hover density across page elements using aggregated user interaction data. The system tracks pixel-level interaction coordinates, normalizes them across viewport sizes and device types, and renders density visualizations where color intensity represents interaction frequency. Supports multiple heatmap types (click, scroll, move) and can segment by user cohort, traffic source, or device type to reveal how different audiences interact with the same page.","intents":["I need to see which buttons and links users actually click vs. where I placed them","I want to understand if users scroll far enough to see my CTA below the fold","I need to compare interaction patterns between mobile and desktop users"],"best_for":["UX designers validating layout assumptions with real user data","Growth teams identifying high-engagement page sections for A/B testing","Content strategists determining optimal CTA placement"],"limitations":["Heatmaps aggregate data and obscure individual user intent; a cluster of clicks may indicate confusion rather than engagement","Viewport normalization across devices can introduce artifacts (e.g., responsive layouts shift element positions)","Requires minimum sample size (~100-500 sessions) to produce statistically meaningful heatmaps; low-traffic pages show noise"],"requires":["UX Sniff tracking script installed","Minimum 100 sessions per heatmap for statistical validity","Page must be publicly accessible (no authentication walls)"],"input_types":["click coordinates (x, y)","scroll depth (pixels or percentage)","hover duration and location","viewport dimensions","device type and OS"],"output_types":["visual heatmap overlay (PNG/SVG)","interaction density JSON (coordinates + frequency)","segmented heatmaps by cohort"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_10","uri":"capability://data.processing.analysis.performance.monitoring.with.page.load.and.interaction.latency.tracking","name":"performance monitoring with page load and interaction latency tracking","description":"Tracks page load performance metrics (time to first byte, first contentful paint, largest contentful paint, cumulative layout shift) and interaction latency (time from user action to visible response) to identify performance-related UX issues. The system correlates performance metrics with user engagement and conversion outcomes to identify if slow pages have higher bounce rates or lower conversion rates. Generates performance reports showing performance variance by device, browser, and geographic region, and alerts when performance degrades below thresholds.","intents":["I need to know if slow page loads are causing users to bounce before they see my content","I want to identify if specific devices or regions experience slower performance and higher drop-off","I need to track if my performance optimizations actually improved user engagement and conversions"],"best_for":["Performance-conscious teams optimizing for conversion and user retention","E-commerce companies where page speed directly impacts revenue","Teams with global users that need to understand regional performance variance"],"limitations":["Performance metrics are collected client-side and vary based on user device, network, and browser; cannot isolate server-side performance issues","Correlation between performance and conversion is probabilistic; slow pages may have low conversion due to other factors (e.g., poor copy, weak CTA)","Performance monitoring adds ~5-10KB to tracking script size and ~1-2% overhead to page load time"],"requires":["UX Sniff tracking script installed","Browser support for Web Vitals API (Chrome 77+, Edge 79+, Safari 15+)","Minimum 500 sessions for statistically meaningful performance analysis"],"input_types":["page load metrics (TTFB, FCP, LCP, CLS)","interaction latency (time from click to response)","device type, browser, network conditions","geographic location"],"output_types":["performance reports (metrics by device, browser, region)","performance vs. conversion correlation analysis","performance alerts (threshold-based notifications)"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_2","uri":"capability://data.processing.analysis.conversion.funnel.analysis.with.drop.off.attribution","name":"conversion funnel analysis with drop-off attribution","description":"Tracks user progression through defined conversion funnels (e.g., landing page → signup → payment) and automatically identifies where users drop off using event-based tracking. The system correlates drop-off events with user attributes (device, traffic source, geography, session duration) and AI-driven behavioral signals to attribute abandonment to specific friction points. Generates reports showing drop-off rates per funnel step, cohort-level conversion variance, and predictive indicators of abandonment (e.g., 'users who hesitate >3 seconds on password field have 60% higher abandonment').","intents":["I need to know exactly which funnel step is losing the most users and why","I want to identify if mobile users drop off at different steps than desktop users","I need to predict which sessions are likely to abandon before they do so I can intervene"],"best_for":["Growth engineers optimizing conversion funnels for SaaS or e-commerce","Product managers prioritizing which funnel step to optimize first","Marketing teams understanding which traffic sources have lowest conversion quality"],"limitations":["Attribution is correlational, not causal; tool cannot determine if a user abandoned due to friction or external factors (e.g., browser crash, network loss)","Requires explicit event tracking setup; if events are misconfigured or missing, funnel analysis becomes unreliable","Predictive abandonment signals are probabilistic and may have false positive rates >20% on low-traffic funnels"],"requires":["UX Sniff tracking script installed","Custom events defined for each funnel step (via API or UI configuration)","Minimum 500 funnel entries per analysis for statistical significance"],"input_types":["custom events (step_viewed, step_completed, step_abandoned)","user attributes (device, source, geography, session_id)","behavioral signals (time_on_step, interaction_count, scroll_depth)"],"output_types":["funnel visualization (step-by-step drop-off rates)","cohort comparison table (conversion rates by segment)","drop-off attribution report (reasons for abandonment)","predictive abandonment scores (0-100 per session)"],"categories":["data-processing-analysis","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_3","uri":"capability://planning.reasoning.ai.generated.ux.insights.and.optimization.recommendations","name":"ai-generated ux insights and optimization recommendations","description":"Analyzes aggregated session, heatmap, and funnel data using machine learning models to identify patterns and generate actionable UX optimization recommendations. The system ingests behavioral data (session replays, interaction heatmaps, conversion funnels, user attributes) and applies pattern-matching algorithms to detect common friction patterns (e.g., 'users consistently hover over button X without clicking', 'form field Y has 40% abandonment rate'). Generates prioritized recommendations with estimated impact (e.g., 'moving CTA above fold could increase conversions by 15%') and links recommendations to supporting evidence (specific sessions, heatmap clusters, funnel drop-off data).","intents":["I don't have time to manually analyze all this data—I need AI to tell me the top 3 things to fix","I want recommendations ranked by estimated impact so I can prioritize my optimization roadmap","I need evidence-backed suggestions I can present to stakeholders, not just hunches"],"best_for":["Product managers and growth leads without dedicated UX research teams","Bootstrapped startups needing data-driven optimization without hiring analysts","Digital agencies delivering UX audit reports to clients"],"limitations":["AI recommendations are pattern-based and may miss context-specific issues (e.g., brand guidelines constraints, technical debt preventing certain changes)","Impact estimates are probabilistic and based on historical data; actual lift may vary 20-50% depending on implementation quality and user base","Recommendations require human interpretation and follow-through; tool does not automate optimizations or A/B test implementations"],"requires":["Minimum 1,000 sessions of data for statistically meaningful recommendations","At least 2 weeks of historical data to establish baseline patterns","Defined conversion goal or success metric"],"input_types":["aggregated session data (count, duration, device, source)","heatmap data (interaction density by element)","funnel data (drop-off rates, cohort variance)","user attributes (device, geography, traffic source)"],"output_types":["prioritized recommendation list (title, description, estimated impact %)","supporting evidence (linked sessions, heatmap screenshots, funnel data)","implementation guidance (specific changes to test)"],"categories":["planning-reasoning","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_4","uri":"capability://data.processing.analysis.real.time.event.tracking.with.custom.event.schema","name":"real-time event tracking with custom event schema","description":"Provides a JavaScript API and UI-based event configuration system for tracking custom user events beyond standard page views and clicks. Developers can define custom events (e.g., 'video_played', 'feature_used', 'error_encountered') with arbitrary properties (event_name, user_id, timestamp, custom_data), then query and segment by those events in dashboards. The system stores events in a time-series database, supports real-time event streaming for live dashboards, and allows retroactive event filtering and segmentation without re-instrumentation.","intents":["I need to track custom product events (e.g., feature usage, error states) beyond standard page analytics","I want to segment users by their custom event history (e.g., 'users who watched tutorial video') for cohort analysis","I need real-time event data to power live dashboards and alerts"],"best_for":["Product teams tracking feature adoption and usage patterns","SaaS companies correlating product usage with churn or expansion revenue","Teams building real-time dashboards or alerts based on user behavior"],"limitations":["Custom event schema requires upfront design; poorly designed events lead to data quality issues and difficult retroactive analysis","Event storage costs scale with event volume; high-frequency events (e.g., every keystroke) can quickly exceed free tier limits","Real-time streaming has ~1-5 second latency; not suitable for sub-second decision-making (e.g., real-time personalization)"],"requires":["UX Sniff tracking script installed","JavaScript knowledge to implement custom event calls (or use UI-based event builder)","Event schema design (naming conventions, property definitions)"],"input_types":["event_name (string)","custom properties (key-value pairs, any JSON type)","user_id (optional, for cross-session tracking)","timestamp (auto-generated or custom)"],"output_types":["event stream (real-time JSON events)","event aggregations (count, unique users, properties)","segmented cohorts (users matching event criteria)"],"categories":["data-processing-analysis","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_5","uri":"capability://data.processing.analysis.cohort.segmentation.and.comparison.with.behavioral.attributes","name":"cohort segmentation and comparison with behavioral attributes","description":"Enables creation of user cohorts based on behavioral attributes (device type, traffic source, geography, session duration, custom events) and compares conversion rates, funnel drop-off, and engagement metrics across cohorts. The system supports both pre-defined cohorts (e.g., 'mobile users', 'organic traffic') and custom cohort definitions using boolean logic (e.g., 'users from US who spent >2 minutes on page AND clicked CTA'). Generates side-by-side comparison reports showing variance in key metrics, statistical significance tests, and cohort-specific heatmaps and session replays.","intents":["I need to understand if mobile users have different conversion rates than desktop users","I want to compare how users from different traffic sources behave on my site","I need to identify high-value user cohorts (e.g., users who engage with feature X) and understand what makes them different"],"best_for":["Growth teams optimizing for specific user segments (mobile, geographic, traffic source)","Product managers understanding feature adoption variance across user types","Marketing teams evaluating traffic quality by source and optimizing spend allocation"],"limitations":["Cohort comparison requires sufficient sample size per cohort (~100+ sessions); small cohorts produce unreliable variance estimates","Statistical significance tests assume independent samples; if cohorts overlap, variance estimates may be biased","Behavioral attributes are limited to tracked data; cannot segment by unmeasured attributes (e.g., user intent, purchase intent)"],"requires":["UX Sniff tracking script installed","Minimum 500 total sessions across all cohorts","Defined conversion goal or success metric for comparison"],"input_types":["behavioral attributes (device, source, geography, session_duration, custom_events)","conversion metrics (funnel completion, engagement signals)","boolean logic for cohort definition (AND, OR, NOT operators)"],"output_types":["cohort comparison table (metrics by cohort with variance)","statistical significance indicators (p-values, confidence intervals)","cohort-specific visualizations (heatmaps, funnel charts, session replays)"],"categories":["data-processing-analysis","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_6","uri":"capability://safety.moderation.privacy.compliant.data.collection.with.configurable.masking","name":"privacy-compliant data collection with configurable masking","description":"Implements privacy-first data collection with configurable PII masking, consent management, and GDPR/CCPA compliance features. The system allows configuration of sensitive data patterns (passwords, credit card numbers, email addresses) to be automatically masked in session replays and event logs. Supports consent-based tracking (opt-in/opt-out), cookie management, and data retention policies. Provides audit logs showing what data was collected, masked, and deleted per user.","intents":["I need to record user sessions for UX analysis without violating GDPR or exposing sensitive data","I want to mask credit card and password fields automatically so my team doesn't see sensitive data","I need to prove to compliance teams that we're handling user data responsibly"],"best_for":["SaaS and fintech companies handling sensitive user data","Teams operating in GDPR/CCPA jurisdictions","Companies with strict data privacy policies or compliance requirements"],"limitations":["Masking is pattern-based and may miss context-specific sensitive data (e.g., custom form fields with PII)","Consent management requires user interaction; cannot track non-consenting users, reducing sample size","Data retention policies are fixed per plan; cannot implement custom retention schedules per data type"],"requires":["UX Sniff tracking script installed","Privacy policy updated to disclose data collection and masking practices","Consent banner or opt-in mechanism on website"],"input_types":["sensitive data patterns (regex or predefined patterns)","user consent signals (cookie, local storage, API)","data retention policy (days/months)"],"output_types":["masked session replays (sensitive fields redacted)","audit logs (data collection, masking, deletion events)","compliance reports (GDPR/CCPA attestations)"],"categories":["safety-moderation","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_7","uri":"capability://tool.use.integration.integration.with.third.party.analytics.and.marketing.platforms","name":"integration with third-party analytics and marketing platforms","description":"Provides integrations with popular analytics, CRM, and marketing automation platforms (e.g., Google Analytics, Segment, Zapier, Slack) to export UX Sniff data and trigger actions based on UX events. The system supports bidirectional data sync (importing user segments from external platforms, exporting UX events to external systems) and webhook-based event streaming for real-time integrations. Enables use cases like 'send Slack notification when funnel drop-off rate exceeds threshold' or 'export high-engagement user cohorts to CRM for targeted outreach'.","intents":["I want to export UX Sniff insights to Google Analytics or Segment for unified analytics","I need to trigger marketing actions (email, SMS, ads) based on UX events (e.g., cart abandonment)","I want to receive Slack alerts when key UX metrics degrade"],"best_for":["Teams using multiple analytics and marketing tools that need unified data flow","Growth teams automating marketing actions based on UX signals","Companies with existing data infrastructure (data warehouse, CDP) that need to integrate UX data"],"limitations":["Integration latency varies by platform; Slack webhooks are near-real-time (~1-5 seconds), but batch exports to data warehouses may have 1-24 hour delays","Requires API keys and authentication setup for each integration; managing credentials across multiple platforms adds operational overhead","Data schema mapping between UX Sniff and external platforms may require custom transformation logic"],"requires":["UX Sniff account with integration tier enabled","API keys or OAuth tokens for target platforms (Google Analytics, Segment, Slack, etc.)","Webhook URL or API endpoint for receiving data"],"input_types":["UX Sniff events (session, heatmap, funnel, custom events)","user cohorts (behavioral segments)","metrics (conversion rates, drop-off rates, engagement signals)"],"output_types":["exported events (JSON, CSV)","webhook payloads (real-time event streams)","CRM/marketing platform actions (email, SMS, ad targeting)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_8","uri":"capability://data.processing.analysis.device.and.browser.compatibility.testing.with.cross.platform.heatmaps","name":"device and browser compatibility testing with cross-platform heatmaps","description":"Automatically segments user sessions and heatmaps by device type (mobile, tablet, desktop), browser (Chrome, Safari, Firefox, Edge), and OS (iOS, Android, Windows, macOS) to identify platform-specific UX issues. The system normalizes interaction coordinates across different viewport sizes and device pixel ratios, then generates device-specific heatmaps and funnel analyses. Enables comparison of conversion rates, engagement metrics, and session replay quality across platforms to identify where users experience friction on specific devices or browsers.","intents":["I need to understand if my site works well on mobile, tablet, and desktop—are conversion rates different?","I want to see heatmaps for mobile users specifically to understand how they interact with my responsive layout","I need to identify if specific browsers (e.g., Safari) have higher bounce rates or technical issues"],"best_for":["E-commerce and SaaS teams optimizing for multi-device user bases","Mobile-first product teams validating responsive design assumptions","QA and product teams identifying browser-specific bugs or compatibility issues"],"limitations":["Viewport normalization across devices can introduce artifacts for responsive layouts that significantly change structure (e.g., hamburger menu on mobile vs. full nav on desktop)","Low traffic on specific device/browser combinations produces unreliable heatmaps; requires minimum ~50-100 sessions per device type","Cannot detect issues that only manifest on specific hardware (e.g., touch latency on older Android devices)"],"requires":["UX Sniff tracking script installed","Minimum 500 total sessions across all device types","Responsive design or separate mobile/desktop versions"],"input_types":["device type (mobile, tablet, desktop)","browser and version (Chrome, Safari, Firefox, Edge)","OS and version (iOS, Android, Windows, macOS)","viewport dimensions and device pixel ratio"],"output_types":["device-specific heatmaps (click, scroll, move)","device-specific funnel analyses (conversion rates by device)","cross-device comparison reports (variance in metrics)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_ux-sniff__cap_9","uri":"capability://data.processing.analysis.geographic.and.traffic.source.segmentation.with.regional.heatmaps","name":"geographic and traffic source segmentation with regional heatmaps","description":"Segments all analytics (sessions, heatmaps, funnels, conversion rates) by geographic location (country, region, city) and traffic source (organic, paid, direct, referral, social) to identify regional and channel-specific UX patterns. The system uses IP geolocation and UTM parameter parsing to assign users to geographic and traffic source cohorts, then generates region-specific and channel-specific heatmaps, funnel analyses, and session replays. Enables identification of regional UX issues (e.g., 'users in Japan have 40% higher drop-off on checkout') or channel-specific problems (e.g., 'paid traffic from Google Ads has lower engagement than organic').","intents":["I need to understand if users from different countries have different conversion rates or UX issues","I want to see if my paid traffic (Google Ads, Facebook) has different behavior than organic traffic","I need to identify if specific traffic sources (e.g., affiliate partners) send low-quality traffic"],"best_for":["Global SaaS and e-commerce companies optimizing for multiple regions","Marketing teams evaluating traffic quality by source and optimizing ad spend","Teams with localized content or regional pricing that need to validate regional UX"],"limitations":["IP geolocation accuracy varies by region; VPN and proxy usage can skew geographic data","UTM parameter tracking requires proper implementation; missing or inconsistent UTM tags reduce traffic source accuracy","Low traffic from specific regions or sources produces unreliable heatmaps; requires minimum ~50-100 sessions per segment"],"requires":["UX Sniff tracking script installed","UTM parameters or traffic source tracking configured","Minimum 500 total sessions across all geographic and traffic source segments"],"input_types":["IP address (for geolocation)","UTM parameters (utm_source, utm_medium, utm_campaign)","referrer header (for organic/referral detection)"],"output_types":["geographic heatmaps (by country, region, city)","traffic source heatmaps (by source, medium, campaign)","regional funnel analyses (conversion rates by geography)","traffic source comparison reports (engagement by source)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":43,"verified":false,"data_access_risk":"high","permissions":["Website with JavaScript enabled","UX Sniff tracking script installed (1-2 lines of code)","GDPR/privacy policy updated to disclose session recording","UX Sniff tracking script installed","Minimum 100 sessions per heatmap for statistical validity","Page must be publicly accessible (no authentication walls)","Browser support for Web Vitals API (Chrome 77+, Edge 79+, Safari 15+)","Minimum 500 sessions for statistically meaningful performance analysis","Custom events defined for each funnel step (via API or UI configuration)","Minimum 500 funnel entries per analysis for statistical significance"],"failure_modes":["Session replay adds ~50-100KB per session to storage; high-traffic sites may hit data retention limits on free tier","AI annotations are pattern-based and may miss context-specific friction (e.g., accessibility issues for screen readers)","Privacy-sensitive data (passwords, credit cards) requires manual masking configuration; no automatic PII detection","Heatmaps aggregate data and obscure individual user intent; a cluster of clicks may indicate confusion rather than engagement","Viewport normalization across devices can introduce artifacts (e.g., responsive layouts shift element positions)","Requires minimum sample size (~100-500 sessions) to produce statistically meaningful heatmaps; low-traffic pages show noise","Performance metrics are collected client-side and vary based on user device, network, and browser; cannot isolate server-side performance issues","Correlation between performance and conversion is probabilistic; slow pages may have low conversion due to other factors (e.g., poor copy, weak CTA)","Performance monitoring adds ~5-10KB to tracking script size and ~1-2% overhead to page load time","Attribution is correlational, not causal; tool cannot determine if a user abandoned due to friction or external factors (e.g., browser crash, network loss)","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.36666666666666664,"quality":0.78,"ecosystem":0.15000000000000002,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.1,"match_graph":0.35,"freshness":0.05}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:33.649Z","last_scraped_at":"2026-04-05T13:23:42.551Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=ux-sniff","compare_url":"https://unfragile.ai/compare?artifact=ux-sniff"}},"signature":"PADSrOK0WNdNdGm+avXbq6i51+FrQnM7S0L4pu1r0o62Y11m8YhiW+BhIEWIvqQEW4NlOoRKV4vlB2cZxnqJCA==","signedAt":"2026-06-21T06:26:13.395Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/ux-sniff","artifact":"https://unfragile.ai/ux-sniff","verify":"https://unfragile.ai/api/v1/verify?slug=ux-sniff","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}