{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"tool_neubird","slug":"neubird","name":"NeuBird","type":"product","url":"https://neubird.ai","page_url":"https://unfragile.ai/neubird","categories":["text-writing"],"tags":[],"pricing":{"model":"paid","free":false,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"tool_neubird__cap_0","uri":"capability://automation.workflow.batch.video.processing.with.parallel.encoding","name":"batch video processing with parallel encoding","description":"Processes multiple video files simultaneously through a distributed encoding pipeline that queues jobs, allocates compute resources dynamically, and manages output coordination across parallel workers. The system likely uses a job queue (Redis/RabbitMQ pattern) to track batch state, distributes encoding tasks across available GPU/CPU resources, and aggregates results into a unified output manifest. This enables creators to submit 10-100+ videos and receive processed outputs without sequential bottlenecks.","intents":["I need to process 50 marketing videos in one batch job without waiting for each to finish sequentially","I want to submit a folder of raw footage and have it all edited and exported overnight","I need to apply the same editing template across dozens of videos for consistency"],"best_for":["marketing teams producing high-volume content (20+ videos/month)","YouTube creators managing weekly upload schedules","social media agencies handling multiple client campaigns simultaneously"],"limitations":["Batch processing latency scales with queue depth — 100-video batches may take 4-8 hours depending on infrastructure","No granular per-video priority queuing — all jobs in a batch treated with equal priority","Output coordination requires external storage (S3/GCS) — local file system batches limited to ~50GB","Concurrent processing limited by account tier; unclear if premium tiers unlock higher parallelism"],"requires":["Video files in MP4, MOV, or WebM format (codec support unknown)","Minimum 100MB free storage for batch output","Active NeuBird account with batch processing tier enabled","Network bandwidth for upload/download (typical 1-5 Mbps sustained)"],"input_types":["video files (MP4, MOV, WebM)","batch manifest (JSON or CSV with file paths and editing parameters)","template configuration (preset editing rules)"],"output_types":["processed video files (MP4, optimized for platform)","batch report (JSON with per-video processing status and metrics)","export links (direct download or cloud storage paths)"],"categories":["automation-workflow","video-processing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_1","uri":"capability://image.visual.automated.silence.detection.and.removal","name":"automated silence detection and removal","description":"Analyzes audio tracks using spectral analysis or ML-based voice activity detection (VAD) to identify silence, filler words, and dead air, then automatically removes or compresses these segments while maintaining audio sync across video tracks. The system likely uses a pre-trained audio classification model (possibly trained on speech/silence patterns) that segments the timeline, marks regions below a configurable threshold, and triggers frame-accurate trimming in the video timeline. This reduces manual scrubbing and cutting work.","intents":["I want to automatically cut out all the pauses and 'umms' from my podcast video without manual editing","I need to tighten up interview footage by removing long silences between Q&A","I want to compress dead air in my stream VOD to make it watchable in half the time"],"best_for":["podcast creators and video bloggers with long-form content","interview-based content producers (journalists, podcasters)","stream VOD editors who need quick turnaround on archive clips"],"limitations":["Silence detection threshold is likely global or preset-based — no per-segment adaptive thresholding, so music/ambient sound may be incorrectly flagged","No speaker diarization — cannot distinguish between multiple speakers or preserve intentional pauses in dialogue","Filler word detection (if present) likely limited to common English patterns; non-English content or accented speech may have poor accuracy","Audio sync drift possible if video contains multiple audio tracks with different silence patterns"],"requires":["Video with clear audio track (mono or stereo; surround sound support unknown)","Minimum audio bitrate 128 kbps for reliable VAD","English language audio (other languages may have degraded performance)"],"input_types":["video file with embedded audio","silence threshold parameter (dB level, typically -40 to -60)","optional: minimum silence duration to remove (e.g., 0.5 seconds)"],"output_types":["trimmed video file with silence segments removed","edit timeline (JSON) showing removed segments with timestamps","audio waveform visualization highlighting detected silence regions"],"categories":["image-visual","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_10","uri":"capability://tool.use.integration.collaborative.editing.with.version.control.and.approval.workflows","name":"collaborative editing with version control and approval workflows","description":"Enables multiple team members to work on the same project with version tracking, commenting, and approval workflows. The system likely implements a centralized project state (stored in cloud database), tracks changes per user with timestamps, supports comment threads on specific timeline segments, and implements approval gates (e.g., 'requires client approval before export'). This enables asynchronous collaboration without file conflicts.","intents":["I need my team to review and approve edits before final export","I want to track who made what changes to the video project","I need to collaborate with a client on edits without sending files back and forth"],"best_for":["agency teams producing client work with approval requirements","distributed teams collaborating across time zones","in-house marketing teams with multiple editors and reviewers"],"limitations":["Real-time collaborative editing likely not supported — changes may require manual sync or refresh","Version history may be limited (e.g., last 10 versions only) — no full audit trail","Commenting is likely limited to timeline segments — no frame-level or pixel-level annotation","Approval workflows are likely simple (approve/reject) — no conditional approval or multi-stage reviews","Concurrent editing by multiple users may cause conflicts — unclear how conflicts are resolved"],"requires":["NeuBird account with collaboration features enabled","Team members with NeuBird accounts","Stable internet connection for real-time sync"],"input_types":["video project (stored in NeuBird cloud)","user permissions (editor, reviewer, approver roles)","optional: approval workflow configuration (required approvers, approval criteria)"],"output_types":["version history (JSON with changes per user and timestamp)","comment thread (text comments with user attribution and timestamps)","approval status report (current approval state and pending approvers)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_11","uri":"capability://planning.reasoning.ai.powered.content.recommendations.and.trending.format.detection","name":"ai-powered content recommendations and trending format detection","description":"Analyzes trending video formats, styles, and content patterns from social media platforms and recommends editing approaches, templates, or content structures that align with current trends. The system likely monitors platform trends (TikTok, YouTube, Instagram) using web scraping or API integration, analyzes successful video characteristics (length, pacing, music, text overlay density), and recommends matching templates or editing parameters. This helps creators stay current with platform trends.","intents":["I want to know what video formats are trending on TikTok right now","I need to edit my video to match current YouTube Shorts trends","I want recommendations on pacing, music, and text overlay based on trending videos"],"best_for":["social media creators wanting to stay current with trends","marketing teams optimizing content for platform algorithms","content creators with limited creative direction seeking inspiration"],"limitations":["Trend detection is likely delayed (trends may be 1-2 weeks old by the time recommendations are generated)","Recommendations are based on aggregate trends — may not account for niche audiences or emerging micro-trends","No causal analysis — cannot determine whether trends are driven by algorithm changes, influencer adoption, or organic user preference","Recommendations may push creators toward generic, trend-chasing content rather than unique creative direction","Trend data may be biased toward high-performing content (survivorship bias) — unsuccessful trend attempts not analyzed"],"requires":["NeuBird account with trend analysis feature enabled","Internet connection for real-time trend data fetching"],"input_types":["target platform (e.g., 'TikTok', 'YouTube Shorts', 'Instagram Reels')","optional: content category (e.g., 'comedy', 'education', 'music')","optional: target audience (e.g., 'Gen Z', 'millennials')"],"output_types":["trend report (JSON with trending formats, styles, and characteristics)","template recommendations (array of templates matching current trends)","editing parameter suggestions (pacing, music style, text overlay density)","trend visualization (charts showing trend growth over time)"],"categories":["planning-reasoning","search-retrieval"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_2","uri":"capability://image.visual.ai.driven.color.grading.and.normalization","name":"ai-driven color grading and normalization","description":"Applies learned color correction profiles to video footage using neural network-based color space transformation, likely trained on professional colorist workflows. The system analyzes frame histograms, detects color casts, and applies LUT (Look-Up Table) transformations or neural color mapping to normalize exposure, saturation, and white balance across clips. This enables consistent color treatment across multi-clip sequences without manual color wheel adjustment.","intents":["I shot footage on different cameras and need consistent color across all clips","I want to apply a cinematic color grade to my raw footage automatically","I need to fix overexposed or underexposed clips without manual adjustment"],"best_for":["multi-camera video producers (interviews, events)","YouTube creators wanting cinematic aesthetics without color grading expertise","marketing teams needing consistent brand color across campaign videos"],"limitations":["Color grading is deterministic and preset-based — no real-time adjustment or A/B preview of different grades","May over-saturate or desaturate footage with unusual lighting (e.g., neon, tungsten) or mixed color temperatures","No per-frame masking or selective color correction — entire clip receives uniform treatment","Output quality depends on source footage quality; heavily compressed or low-bitrate sources may show banding artifacts"],"requires":["Video file with color information (8-bit or 10-bit color depth preferred)","Footage shot in standard color space (Rec.709 or DCI-P3; log footage support unknown)"],"input_types":["video file (MP4, MOV, ProRes)","color grade preset name (e.g., 'cinematic', 'warm', 'cool')","optional: intensity slider (0-100) to control grade strength"],"output_types":["color-graded video file","LUT file (if exportable) for use in other editors","before/after comparison image"],"categories":["image-visual","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_3","uri":"capability://image.visual.intelligent.clip.segmentation.and.scene.detection","name":"intelligent clip segmentation and scene detection","description":"Analyzes video content using computer vision (shot boundary detection, scene change detection) and audio cues (dialogue, music transitions) to automatically segment footage into logical clips. The system likely uses frame-to-frame optical flow analysis or neural scene classification to detect cuts, camera movements, and content changes, then creates edit points at natural boundaries. This enables automatic clip organization without manual timeline scrubbing.","intents":["I have 2 hours of raw footage and need it automatically split into usable clips","I want to detect where each scene or topic change happens in my interview","I need to organize my footage into chapters automatically for easier editing"],"best_for":["documentary and interview editors processing long-form raw footage","podcast video producers converting audio content to video","content creators with large archives needing rapid organization"],"limitations":["Scene detection may fail on slow transitions (fades, dissolves) or static shots with subtle content changes","No semantic understanding of content — cannot distinguish between intentional scene breaks and accidental camera movements","Audio-based segmentation (dialogue/music detection) likely limited to clear, high-quality audio; background noise may cause false positives","Clip boundaries may not align with creative intent — requires manual review and adjustment"],"requires":["Video with clear visual or audio transitions (minimum 720p resolution for reliable detection)","Audio track for dialogue/music detection (mono or stereo)"],"input_types":["video file (MP4, MOV, WebM)","optional: sensitivity parameter (low/medium/high) for scene detection threshold","optional: minimum clip duration (e.g., 5 seconds minimum per clip)"],"output_types":["clip manifest (JSON with timestamps and suggested clip names)","segmented video files (one per detected scene)","timeline visualization showing detected boundaries"],"categories":["image-visual","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_4","uri":"capability://automation.workflow.template.based.editing.workflow.with.preset.rules","name":"template-based editing workflow with preset rules","description":"Provides pre-configured editing templates that encode common workflows (e.g., 'YouTube intro + body + outro', 'Instagram Reel format', 'podcast thumbnail + clips') as rule sets that automatically apply transitions, text overlays, music, and export settings. Templates likely store editing parameters as JSON/YAML configurations that the system applies sequentially to input footage, with variable substitution for titles, dates, and branding elements. This enables one-click application of complex editing sequences.","intents":["I want to apply my brand's standard editing style to all my videos automatically","I need to create YouTube videos with consistent intro/outro/branding across all uploads","I want to generate Instagram Reels with the right aspect ratio, music, and text automatically"],"best_for":["marketing teams with brand guidelines requiring consistent editing","YouTube creators producing weekly content with repeating structure","social media managers handling multiple accounts with platform-specific formats"],"limitations":["Templates are rigid — limited customization without creating new templates or manual post-processing","No template versioning or A/B testing — cannot compare different template outputs","Music/audio in templates may have licensing restrictions or limited library","Text overlay positioning and styling may not adapt to different video aspect ratios or content length"],"requires":["NeuBird account with template library access","Video footage matching template assumptions (e.g., 'YouTube template' expects 16:9 footage)","Optional: brand assets (logos, color palettes) for customization"],"input_types":["video file (MP4, MOV)","template name (string identifier)","optional: template variables (JSON with title, date, brand colors, etc.)","optional: music/audio file for template audio track"],"output_types":["edited video file with template applied","template configuration file (for reuse or modification)","export presets (platform-specific settings for YouTube, Instagram, TikTok)"],"categories":["automation-workflow","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_5","uri":"capability://automation.workflow.multi.platform.export.optimization.with.format.conversion","name":"multi-platform export optimization with format conversion","description":"Automatically generates platform-optimized video exports (YouTube, Instagram, TikTok, LinkedIn, etc.) with correct aspect ratios, bitrates, codecs, and metadata. The system likely maintains a database of platform specifications (resolution, frame rate, duration limits, safe area margins) and applies appropriate encoding parameters, watermark placement, and subtitle formatting per platform. This eliminates manual re-encoding and format conversion work.","intents":["I need to export my video for YouTube, Instagram, and TikTok with correct specs automatically","I want to generate multiple aspect ratio versions (16:9, 9:16, 1:1) from a single edit","I need to add platform-specific subtitles and safe area margins without manual adjustment"],"best_for":["social media managers distributing content across multiple platforms","YouTube creators also publishing to Instagram/TikTok","marketing teams needing rapid multi-platform campaign deployment"],"limitations":["Platform specifications change frequently — export presets may become outdated without regular updates","No custom platform support — limited to pre-configured platforms (YouTube, Instagram, TikTok, LinkedIn, etc.)","Aspect ratio conversion (e.g., 16:9 to 9:16) may require pillarboxing/letterboxing or content reframing, which may crop important elements","Watermark placement is likely fixed — no per-platform watermark customization","Subtitle formatting may not support all platform-specific requirements (e.g., TikTok's caption styling)"],"requires":["Video file in standard format (MP4, MOV)","Platform accounts (YouTube, Instagram, etc.) if direct upload is supported","Sufficient storage for multiple export versions (typically 2-5x source file size)"],"input_types":["edited video file","list of target platforms (array of platform names)","optional: custom metadata (title, description, tags per platform)"],"output_types":["platform-optimized video files (one per platform)","metadata files (JSON with platform-specific descriptions, hashtags, etc.)","upload manifest (instructions for uploading to each platform)"],"categories":["automation-workflow","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_6","uri":"capability://data.processing.analysis.ai.generated.captions.and.subtitle.generation","name":"ai-generated captions and subtitle generation","description":"Automatically transcribes video audio using speech-to-text (likely a pre-trained ASR model such as Whisper or similar) and generates synchronized subtitle files (SRT, VTT, or hardcoded burnt-in captions). The system performs audio-to-text conversion, timestamps each subtitle segment to match speech timing, and optionally applies speaker diarization to label different speakers. This eliminates manual transcription and caption timing work.","intents":["I need to generate captions for my video automatically without manual transcription","I want to create multi-language subtitles from English audio","I need to identify who is speaking in my interview and label speakers in captions"],"best_for":["content creators producing accessibility-compliant videos","international creators needing multi-language subtitles","interview/podcast producers needing speaker identification"],"limitations":["ASR accuracy depends on audio quality — background noise, accents, and technical jargon may cause transcription errors","Speaker diarization (if supported) likely limited to 2-3 speakers; complex multi-speaker scenarios may fail","No context awareness — homophones and domain-specific terms may be transcribed incorrectly (e.g., 'their' vs 'there')","Multi-language support likely limited to major languages; rare languages may not be supported","Subtitle timing may drift on long videos or videos with variable speech pace"],"requires":["Video with clear audio track (mono or stereo; surround sound support unknown)","Audio bitrate minimum 128 kbps for reliable transcription","English language audio (other languages may have degraded accuracy)"],"input_types":["video file with audio track","optional: language code (e.g., 'en', 'es', 'fr')","optional: speaker names (for diarization labeling)"],"output_types":["subtitle file (SRT, VTT, or ASS format)","transcript text file (plain text or JSON with timestamps)","hardcoded caption video (if burnt-in captions requested)","speaker diarization report (JSON with speaker labels and timestamps)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_7","uri":"capability://image.visual.dynamic.text.overlay.and.title.generation","name":"dynamic text overlay and title generation","description":"Automatically generates and positions text overlays (titles, lower thirds, captions, call-to-action text) on video frames using layout templates and content-aware placement. The system likely analyzes frame composition (using object detection or safe area analysis) to position text in non-intrusive locations, applies typography presets (fonts, colors, animations), and synchronizes text timing with audio or scene changes. This enables rapid text addition without manual keyframing.","intents":["I want to add animated titles and lower thirds to my video automatically","I need to place call-to-action text at the end of my video without manual positioning","I want to add chapter titles that sync with scene changes automatically"],"best_for":["YouTube creators needing rapid title/CTA addition","marketing teams creating templated promotional videos","social media creators adding text overlays for engagement"],"limitations":["Text placement is likely rule-based (e.g., 'top-left', 'center', 'bottom-right') — no pixel-level control or custom positioning","Content-aware placement may fail on complex scenes with multiple subjects or text-heavy backgrounds","Font and color options likely limited to preset palettes — no custom font upload or color picker","Text animation options may be limited to simple effects (fade, slide) — no complex keyframe animations","No text wrapping or dynamic sizing based on content length — long text may overflow or be truncated"],"requires":["Video file (MP4, MOV)","Text content (string or array of text segments)"],"input_types":["video file","text content (string or JSON array with text, timing, and position)","optional: typography preset (font, color, size, animation)","optional: position hint (e.g., 'top-left', 'center', 'bottom-right')"],"output_types":["video with text overlays applied","text overlay configuration file (for reuse or modification)","preview image showing text placement"],"categories":["image-visual","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_8","uri":"capability://automation.workflow.background.music.and.sound.design.library.integration","name":"background music and sound design library integration","description":"Provides access to a curated library of royalty-free background music, sound effects, and ambient audio tracks that can be automatically matched to video content and inserted at appropriate timing. The system likely analyzes video mood/pace (using scene detection or metadata) and recommends matching audio tracks, then synchronizes audio levels and applies crossfades to blend with existing audio. This eliminates manual music search and audio mixing work.","intents":["I need to add background music to my video without copyright issues","I want to find sound effects that match my video's mood automatically","I need to layer multiple audio tracks with proper mixing automatically"],"best_for":["content creators needing quick audio addition without licensing concerns","marketing teams creating branded videos with consistent audio","YouTube creators avoiding copyright strikes on monetized content"],"limitations":["Music library is likely limited in size and genre diversity compared to premium services (Epidemic Sound, Artlist)","Automatic music matching may not align with creative intent — recommendations based on metadata rather than subjective mood","Audio mixing is likely automatic (volume normalization, crossfades) — no manual mixing control","Licensing terms may restrict commercial use or require attribution","No custom audio upload — limited to library tracks"],"requires":["NeuBird account with music library access","Video file with clear audio track (for level matching)"],"input_types":["video file","optional: mood/genre preference (string or array)","optional: duration requirement (for music track length)"],"output_types":["video with background music added","audio track file (for separate use)","music metadata (title, artist, license terms)"],"categories":["automation-workflow","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_neubird__cap_9","uri":"capability://data.processing.analysis.video.quality.analysis.and.optimization.recommendations","name":"video quality analysis and optimization recommendations","description":"Analyzes source footage for technical quality issues (exposure, focus, color cast, compression artifacts, frame rate inconsistencies) and provides optimization recommendations or automatic corrections. The system likely uses computer vision analysis (histogram analysis, blur detection, color space analysis) to identify problems and suggests corrections (e.g., 'increase exposure by 1.5 stops', 'apply denoising filter'). This helps creators identify and fix footage issues before editing.","intents":["I want to know if my footage is good enough quality before spending time editing","I need to fix overexposed or underexposed clips automatically","I want to identify and remove blurry or out-of-focus clips from my batch"],"best_for":["content creators with limited technical knowledge assessing footage quality","batch processors needing to filter low-quality clips before processing","videographers reviewing footage from multiple cameras for consistency"],"limitations":["Quality analysis is objective (technical metrics) — does not account for creative intent or artistic choices","Blur detection may flag intentional motion blur or shallow depth-of-field as defects","Exposure and color analysis may not account for intentional color grading or stylistic choices","Recommendations are suggestions only — automatic corrections may degrade footage if applied incorrectly","No semantic understanding of content — cannot assess whether footage is usable based on content (e.g., identifying unusable takes)"],"requires":["Video file (MP4, MOV, ProRes)","Minimum resolution 720p for reliable analysis"],"input_types":["video file","optional: quality threshold (e.g., 'minimum acceptable exposure level')"],"output_types":["quality report (JSON with detected issues and severity levels)","optimization recommendations (array of suggested corrections with parameters)","quality score (0-100 rating)","before/after preview images (if corrections applied)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":43,"verified":false,"data_access_risk":"high","permissions":["Video files in MP4, MOV, or WebM format (codec support unknown)","Minimum 100MB free storage for batch output","Active NeuBird account with batch processing tier enabled","Network bandwidth for upload/download (typical 1-5 Mbps sustained)","Video with clear audio track (mono or stereo; surround sound support unknown)","Minimum audio bitrate 128 kbps for reliable VAD","English language audio (other languages may have degraded performance)","NeuBird account with collaboration features enabled","Team members with NeuBird accounts","Stable internet connection for real-time sync"],"failure_modes":["Batch processing latency scales with queue depth — 100-video batches may take 4-8 hours depending on infrastructure","No granular per-video priority queuing — all jobs in a batch treated with equal priority","Output coordination requires external storage (S3/GCS) — local file system batches limited to ~50GB","Concurrent processing limited by account tier; unclear if premium tiers unlock higher parallelism","Silence detection threshold is likely global or preset-based — no per-segment adaptive thresholding, so music/ambient sound may be incorrectly flagged","No speaker diarization — cannot distinguish between multiple speakers or preserve intentional pauses in dialogue","Filler word detection (if present) likely limited to common English patterns; non-English content or accented speech may have poor accuracy","Audio sync drift possible if video contains multiple audio tracks with different silence patterns","Real-time collaborative editing likely not supported — changes may require manual sync or refresh","Version history may be limited (e.g., last 10 versions only) — no full audit trail","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.36666666666666664,"quality":0.78,"ecosystem":0.2,"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":"2026-05-24T12:16:31.858Z","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=neubird","compare_url":"https://unfragile.ai/compare?artifact=neubird"}},"signature":"/dyCbZP4jLVhzhHWf1Nz3s3/2h8rI2ltitdjo1EoHic9VpBWhlQlFSmimKFi0tPxRxrD9vZp228/txZXzl7oAA==","signedAt":"2026-06-21T07:51:00.892Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/neubird","artifact":"https://unfragile.ai/neubird","verify":"https://unfragile.ai/api/v1/verify?slug=neubird","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"}}