{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"tool_workhub","slug":"workhub","name":"WorkHub","type":"product","url":"https://www.workhub.ai","page_url":"https://unfragile.ai/workhub","categories":["automation"],"tags":[],"pricing":{"model":"freemium","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"tool_workhub__cap_0","uri":"capability://memory.knowledge.privacy.first.knowledge.consolidation.with.local.llm.inference","name":"privacy-first knowledge consolidation with local llm inference","description":"WorkHub consolidates dispersed organizational knowledge (documents, chat logs, databases) into a unified searchable index while performing AI analysis using on-premise or edge-deployed language models rather than sending data to third-party cloud AI providers. This architecture keeps sensitive data within organizational boundaries during both indexing and inference phases, using local embedding models and retrieval-augmented generation (RAG) pipelines that never expose raw content to external APIs.","intents":["I need to build a knowledge base from internal documents without exposing proprietary data to OpenAI or other cloud LLM providers","Our compliance team requires that all AI processing stays within our infrastructure or a private cloud deployment","I want to consolidate knowledge from multiple sources (Slack, email, wikis, databases) while maintaining data residency requirements"],"best_for":["Healthcare organizations processing PHI and bound by HIPAA","Financial services firms managing PII and regulatory data","Government agencies with data classification requirements","Enterprises in EU/APAC with GDPR/data localization mandates"],"limitations":["Local LLM inference typically 2-5x slower than cloud APIs (GPT-4) due to hardware constraints; inference latency scales with model size","Requires dedicated compute infrastructure (GPU/TPU) for reasonable performance; no serverless option for variable workloads","Knowledge consolidation from unstructured sources requires manual schema definition; no automatic format detection across heterogeneous data types","Privacy guarantees only as strong as underlying infrastructure security; misconfiguration of network isolation can expose data"],"requires":["On-premise server infrastructure or private cloud account (AWS PrivateLink, Azure ExpressRoute, or equivalent)","GPU compute (NVIDIA A100 or equivalent) for sub-second inference on models >7B parameters","Data connectors for source systems (Slack API, email server access, database credentials)","Network isolation from public internet or VPN tunnel to WorkHub infrastructure"],"input_types":["unstructured text (documents, chat logs, email)","structured data (database tables, CSV, JSON)","semi-structured (PDFs, web pages, wiki markup)"],"output_types":["vector embeddings (stored in local vector database)","indexed knowledge graph","retrieval results with source attribution"],"categories":["memory-knowledge","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_1","uri":"capability://data.processing.analysis.automated.knowledge.extraction.and.schema.mapping.from.heterogeneous.sources","name":"automated knowledge extraction and schema mapping from heterogeneous sources","description":"WorkHub automatically ingests data from multiple source systems (databases, APIs, file storage, communication platforms) and maps unstructured content to a unified knowledge schema using local LLM-based extraction without manual field mapping. The system learns schema patterns from sample documents and applies extraction rules across new incoming data, handling format variations and incomplete fields gracefully.","intents":["I have customer data spread across Salesforce, Slack, email, and internal wikis—I need to unify it into one searchable knowledge base without manual ETL","Our documentation is inconsistently formatted across teams; I want AI to normalize it into a standard schema automatically","I need to extract structured metadata (project status, owner, deadline) from unstructured meeting notes and emails at scale"],"best_for":["Mid-market enterprises with 5+ disconnected data sources and no dedicated data engineering team","Organizations with high documentation churn where manual schema maintenance is infeasible","Teams managing customer/project data across multiple systems of record"],"limitations":["Schema inference accuracy depends on sample quality; ambiguous or sparse examples lead to incorrect field mappings requiring manual correction","Extraction latency scales with document complexity; dense PDFs or images require OCR preprocessing adding 500ms-2s per document","No built-in conflict resolution for duplicate or contradictory data across sources; requires manual deduplication rules","Extraction rules are not portable across schema versions; schema changes require retraining on new samples"],"requires":["API credentials or database access for each source system","Sample documents (5-10 per schema variant) for the system to learn extraction patterns","Defined target schema (JSON schema or equivalent) for normalized output","Ongoing manual review of extraction results for quality assurance (recommended 5-10% sampling)"],"input_types":["structured data (SQL databases, REST APIs)","semi-structured (JSON, XML, CSV)","unstructured text (email, chat logs, documents)","images (PDFs, scanned documents with OCR)"],"output_types":["normalized JSON documents conforming to target schema","extraction confidence scores per field","audit logs showing source and extraction timestamp"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_10","uri":"capability://text.generation.language.intelligent.document.summarization.and.key.insight.extraction","name":"intelligent document summarization and key insight extraction","description":"WorkHub automatically generates summaries of long documents and extracts key insights (decisions, action items, risks, stakeholders) using local LLM inference. Summaries are customizable by length and focus (executive summary, technical details, action items), and extracted insights are indexed separately for quick retrieval without reading full documents.","intents":["I have a 50-page project report but only need the key decisions and action items—I want an AI-generated summary","I want to extract all risks mentioned in a document so I can track them in our risk register","I need to quickly understand what a document is about without reading it in full"],"best_for":["Organizations with large volumes of long-form documentation (reports, meeting notes, proposals)","Teams needing to extract structured insights from unstructured documents","Knowledge managers maintaining searchable summaries for quick reference"],"limitations":["Summarization quality degrades for documents >10k tokens; context window limitations force truncation or hierarchical summarization","Extracted insights are only as good as the LLM's understanding; domain-specific insights may be missed or misinterpreted","Summaries are static; if the source document is updated, summaries are not automatically refreshed","No user feedback loop to correct or improve summaries; poor summaries require manual regeneration"],"requires":["Local LLM model with sufficient context window (2k+ tokens for reasonable document length)","Document content in text format (PDFs require OCR preprocessing)","Summary template or prompt defining desired output format","Storage for generated summaries and extracted insights"],"input_types":["document content (text)","summary style preference (executive, technical, action items)"],"output_types":["generated summary (variable length)","extracted key insights (structured data)","confidence scores for extracted information"],"categories":["text-generation-language","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_11","uri":"capability://search.retrieval.federated.search.across.multiple.knowledge.bases.with.result.ranking","name":"federated search across multiple knowledge bases with result ranking","description":"WorkHub enables searching across multiple independent knowledge bases (e.g., different departments, projects, or organizations) in a single query, with results ranked by relevance and source. The system handles schema differences between knowledge bases, deduplicates results, and provides source attribution so users understand which knowledge base each result came from.","intents":["I want to search across our engineering, product, and operations knowledge bases in one query without switching between systems","I need to find similar issues reported in different departments' knowledge bases to identify patterns","I want to search across partner or customer knowledge bases while maintaining data isolation"],"best_for":["Large enterprises with multiple departments or business units maintaining separate knowledge bases","Organizations with federated data governance where different teams control their own data","Multi-tenant platforms where customers have isolated knowledge bases but need cross-tenant search"],"limitations":["Federated search latency increases with number of knowledge bases; searching 10+ bases can add 2-5s latency due to parallel queries","Result deduplication is imperfect; similar documents from different bases may not be recognized as duplicates","Ranking across heterogeneous sources is challenging; relevance scores from different embedding models are not directly comparable","Cross-base search requires network connectivity to all knowledge bases; offline bases are excluded from results"],"requires":["Multiple knowledge bases with vector embeddings and metadata","Federated search coordinator (central service or distributed protocol)","Result ranking algorithm that normalizes scores across sources","Network connectivity between search coordinator and all knowledge bases"],"input_types":["natural language query","optional filters (source knowledge base, date range, document type)"],"output_types":["ranked results from multiple knowledge bases","source attribution per result","relevance scores normalized across sources"],"categories":["search-retrieval","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_2","uri":"capability://search.retrieval.ai.powered.semantic.search.across.consolidated.knowledge.base","name":"ai-powered semantic search across consolidated knowledge base","description":"WorkHub indexes all consolidated knowledge using vector embeddings generated by local embedding models, enabling semantic search that understands intent and context rather than keyword matching. Queries are embedded in the same vector space as documents, and the system returns ranked results based on semantic similarity with optional filtering by metadata, source system, or recency.","intents":["I want to search our knowledge base by meaning ('How do we handle refunds?') rather than exact keywords, across all sources","I need to find similar past customer issues or project decisions without knowing the exact terminology used","I want search results ranked by relevance to my specific context, not just keyword frequency"],"best_for":["Support teams searching for solutions across fragmented knowledge bases","Product teams finding precedents and design decisions from past projects","Compliance teams searching for policy references across documentation"],"limitations":["Embedding quality depends on model choice; smaller models (384-dim) miss nuanced semantic distinctions vs. larger models (1536-dim) which require more compute","Semantic search can return false positives for polysemous queries ('bank' as financial institution vs. riverbank); requires query clarification or multi-stage ranking","No built-in handling of domain-specific terminology; generic embeddings may conflate industry jargon with common language","Search latency increases with knowledge base size; >1M documents require approximate nearest neighbor search (HNSW) with 5-10% recall loss"],"requires":["Consolidated knowledge base with vector embeddings pre-computed","Vector database (Milvus, Weaviate, or proprietary) with HNSW or IVF indexing","Local embedding model (e.g., sentence-transformers) compatible with document domain","Query interface (API or UI) for end users"],"input_types":["natural language queries (text)","optional metadata filters (source, date range, document type)"],"output_types":["ranked list of documents with similarity scores","highlighted passages matching query intent","source attribution and metadata"],"categories":["search-retrieval","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_3","uri":"capability://automation.workflow.automated.workflow.orchestration.for.knowledge.maintenance.and.data.synchronization","name":"automated workflow orchestration for knowledge maintenance and data synchronization","description":"WorkHub automates repetitive data management tasks—syncing knowledge base updates from source systems, triggering document reviews when content ages, notifying teams of schema violations, and executing multi-step workflows (extract → normalize → validate → publish) without manual intervention. Workflows are defined declaratively using a condition-action model and execute on schedules or event triggers.","intents":["I want to automatically sync our Salesforce customer data into the knowledge base every 4 hours without manual exports","I need to flag documentation older than 6 months for review and notify the owning team automatically","I want to validate that all new documents conform to our schema before they're searchable, and reject non-conforming entries"],"best_for":["Teams managing high-volume data ingestion from multiple sources","Organizations with strict data governance requiring automated compliance checks","Knowledge managers needing to maintain freshness without dedicated data ops staff"],"limitations":["Workflow execution is sequential by default; parallel execution requires explicit configuration and adds complexity","Error handling is basic—failed steps halt the workflow; no built-in retry logic or dead-letter queues for failed records","Workflow monitoring and debugging are limited; no detailed execution traces or performance metrics per step","Scaling to >10k documents per sync cycle may require manual optimization of extraction and validation steps"],"requires":["Defined workflow schema (trigger conditions, action steps, error handlers)","Source system credentials and API access for data extraction","Target schema and validation rules for data conformance","Notification channels (email, Slack, webhook) for alerts and status updates"],"input_types":["workflow definitions (JSON or YAML)","schedule expressions (cron syntax)","event triggers (API webhook, database change, file upload)"],"output_types":["execution logs with step-by-step status","notifications to configured channels","updated knowledge base with synced/validated data"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_4","uri":"capability://text.generation.language.context.aware.ai.chat.interface.with.knowledge.base.grounding","name":"context-aware ai chat interface with knowledge base grounding","description":"WorkHub provides a conversational interface where users query the consolidated knowledge base through natural language. The chat system retrieves relevant documents using semantic search, grounds responses in retrieved content (preventing hallucination), and maintains conversation context across multiple turns. Responses include source citations and confidence scores, enabling users to verify information.","intents":["I want to ask questions about our internal policies and get answers grounded in actual documentation, not AI hallucinations","I need a conversational interface that remembers context from previous questions in the same session","I want to know where information came from so I can verify it or find more details in the source document"],"best_for":["Support teams answering customer questions using internal knowledge","Employees onboarding and learning company policies through conversational interface","Compliance teams verifying policy adherence with cited sources"],"limitations":["Grounding quality depends on retrieval quality; if semantic search misses relevant documents, the LLM cannot answer accurately","Multi-turn context is limited by token budget; conversations >10 turns may lose early context or require summarization","Citation accuracy is not guaranteed; LLM may cite documents that don't actually support the claim (citation hallucination)","No built-in user feedback loop; incorrect answers are not automatically flagged for retraining or correction"],"requires":["Populated knowledge base with vector embeddings","Local LLM model (7B+ parameters for coherent multi-turn conversation)","Semantic search backend for document retrieval","Conversation state management (session storage, context summarization)"],"input_types":["natural language questions (text)","optional conversation history for context"],"output_types":["natural language response","source citations with document IDs and passages","confidence/relevance scores"],"categories":["text-generation-language","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_5","uri":"capability://safety.moderation.role.based.access.control.and.data.visibility.filtering","name":"role-based access control and data visibility filtering","description":"WorkHub enforces fine-grained access control at the document and field level based on user roles and attributes. When a user searches or queries the knowledge base, results are filtered to show only documents they have permission to access. Field-level filtering redacts sensitive information (e.g., salary data, customer PII) based on user role, even within documents the user can access.","intents":["I need to ensure support staff can only see customer data relevant to their assigned accounts, not all customer information","Our finance team should see budget documents, but not individual salary information within those documents","I want to grant contractors access to project documentation while excluding confidential strategic plans"],"best_for":["Enterprises with strict data compartmentalization requirements","Organizations with mixed internal/external users (employees, contractors, partners)","Regulated industries requiring audit trails of who accessed what data"],"limitations":["Role definitions must be maintained manually or synced from external identity provider (Okta, Azure AD); no automatic role inference","Field-level filtering adds latency to search and retrieval; filtering >100 fields per document can add 100-500ms per query","No dynamic attribute-based access control (ABAC) based on document content; rules are static per role","Audit logging of access is available but requires manual review; no built-in anomaly detection for suspicious access patterns"],"requires":["Identity provider integration (LDAP, OAuth, SAML) or manual user/role management","Role definitions with associated document and field permissions","Audit logging infrastructure for compliance reporting","Regular access review process to maintain role accuracy"],"input_types":["user identity and role attributes","document metadata and classification tags","field-level sensitivity labels"],"output_types":["filtered search results visible to user","redacted document content with sensitive fields removed","audit logs of access events"],"categories":["safety-moderation","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_6","uri":"capability://data.processing.analysis.document.classification.and.metadata.tagging.with.llm.based.auto.labeling","name":"document classification and metadata tagging with llm-based auto-labeling","description":"WorkHub automatically classifies documents and assigns metadata tags using local LLM inference based on document content and predefined classification schemas. Users can define custom taxonomies (e.g., document type, project, priority, sensitivity level), and the system applies labels automatically during ingestion. Manual corrections feed back into the classification model to improve accuracy over time.","intents":["I want to automatically tag all incoming documents with their type (policy, procedure, template, etc.) without manual review","I need to classify customer issues by severity and category so they route to the right team automatically","I want to label all documents containing PII or financial data so they're handled according to compliance rules"],"best_for":["Organizations ingesting high-volume unstructured documents requiring consistent categorization","Teams needing to enforce data governance rules based on document classification","Knowledge managers maintaining taxonomies across large, heterogeneous document collections"],"limitations":["Classification accuracy depends on training examples; <10 examples per class leads to poor generalization","Multi-label classification (documents with multiple tags) is less accurate than single-label due to label correlation complexity","Hierarchical taxonomies (parent-child classifications) require explicit training; flat taxonomies are more reliable","Retraining on corrected labels requires manual triggering; no continuous online learning"],"requires":["Predefined classification schema or taxonomy (JSON schema with allowed values)","Training examples (5-20 per class) for the LLM to learn classification patterns","Feedback mechanism for users to correct misclassifications","Periodic retraining pipeline to incorporate corrections"],"input_types":["document content (text, structured metadata)","predefined classification schema"],"output_types":["assigned tags/labels with confidence scores","classification audit trail showing reasoning"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_7","uri":"capability://safety.moderation.compliance.monitoring.and.policy.violation.detection","name":"compliance monitoring and policy violation detection","description":"WorkHub continuously monitors the knowledge base for compliance violations—documents containing sensitive data without proper classification, outdated policies still marked as current, unauthorized data access patterns, or content violating regulatory requirements. The system uses local LLM-based pattern matching and rule engines to flag violations and notify compliance teams with remediation recommendations.","intents":["I need to detect if any documents contain unredacted PII or financial data that should be classified as sensitive","I want to be alerted if someone tries to access customer data outside their assigned accounts","I need to ensure all policies are reviewed and updated within 12 months, with automatic flagging of stale content"],"best_for":["Regulated industries (healthcare, finance, government) with mandatory compliance audits","Organizations with data residency or data protection requirements (GDPR, HIPAA, SOC 2)","Enterprises managing sensitive customer or employee data requiring continuous monitoring"],"limitations":["Pattern detection for sensitive data (PII, PHI, financial info) relies on regex and keyword matching; context-aware detection requires manual rule definition","False positive rate can be high for generic patterns (email addresses, phone numbers) without domain-specific tuning","Compliance rules are static; no automatic adaptation to new regulatory requirements","Remediation recommendations are generic; no integration with ticketing systems for automated remediation workflows"],"requires":["Defined compliance rules and policies (regulatory requirements, internal standards)","Patterns or keywords for sensitive data detection","Audit logging infrastructure with queryable access logs","Notification channels (email, Slack, SIEM) for alerts","Compliance team to review and act on violations"],"input_types":["compliance rule definitions","sensitive data patterns (regex, keyword lists)","access logs and audit trails"],"output_types":["compliance violation alerts with severity","affected documents and remediation recommendations","compliance audit reports"],"categories":["safety-moderation","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_8","uri":"capability://data.processing.analysis.multi.source.data.synchronization.with.conflict.resolution","name":"multi-source data synchronization with conflict resolution","description":"WorkHub maintains synchronization between the consolidated knowledge base and multiple source systems (Salesforce, databases, file storage, APIs) using change detection and conflict resolution strategies. When data changes in a source system, WorkHub detects the change, applies transformations, and updates the knowledge base. If the same data is modified in both the source and knowledge base, the system applies a configured conflict resolution strategy (last-write-wins, source-of-truth, manual review).","intents":["I want to keep our knowledge base in sync with Salesforce so customer information is always current without manual exports","If a document is edited both in our wiki and in the knowledge base, I need a clear strategy for which version wins","I need to detect when data in the source system has changed and automatically update the knowledge base"],"best_for":["Organizations with multiple systems of record requiring unified views","Teams managing data that changes frequently in source systems","Enterprises needing to maintain data consistency across platforms"],"limitations":["Change detection latency depends on source system capabilities; systems without change data capture (CDC) require polling, adding 5-60 minute delays","Conflict resolution strategies are predefined; no intelligent merging of conflicting changes (e.g., merging edits to different fields)","Bi-directional sync is not supported; changes flow only from source to knowledge base or vice versa, not both","Large-scale syncs (>100k records) may require manual batching and scheduling to avoid performance impact"],"requires":["API access or database credentials for each source system","Change detection mechanism (CDC, webhooks, polling, or manual triggers)","Conflict resolution strategy definition (last-write-wins, source-of-truth, manual review)","Transformation rules mapping source schema to knowledge base schema","Monitoring and alerting for sync failures"],"input_types":["change events from source systems (API webhooks, database logs)","source and target data schemas","conflict resolution rules"],"output_types":["synchronized records in knowledge base","sync status and error logs","conflict alerts requiring manual resolution"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"tool_workhub__cap_9","uri":"capability://automation.workflow.document.versioning.and.change.tracking.with.audit.trails","name":"document versioning and change tracking with audit trails","description":"WorkHub maintains a complete version history of all documents in the knowledge base, tracking who changed what, when, and why. Users can view document versions, compare changes between versions (diff), revert to previous versions, and see an audit trail of all modifications. Version history is immutable and tamper-proof for compliance purposes.","intents":["I need to see who edited a policy document and what changed so I can understand the evolution of our procedures","I want to revert a document to a previous version if a recent edit introduced errors","I need an immutable audit trail of all document changes for compliance and regulatory audits"],"best_for":["Organizations with strict compliance and audit requirements","Teams collaborating on shared documents where change tracking is critical","Knowledge managers needing to understand document evolution and maintain quality"],"limitations":["Version storage scales linearly with document size and change frequency; high-churn documents can consume significant storage","Diff computation for large documents (>10MB) can be slow; binary documents (images, PDFs) don't support meaningful diffs","Version history is not automatically cleaned up; retention policies must be manually configured to manage storage","Reverting to a previous version doesn't automatically notify dependent documents or downstream systems"],"requires":["Version storage backend (database or object storage) with sufficient capacity","User identity tracking for audit attribution","Diff algorithm implementation (line-based for text, block-based for binary)","Retention policy configuration for version cleanup"],"input_types":["document content (text, binary)","change metadata (user, timestamp, change reason)"],"output_types":["version history with metadata","diffs between versions","audit trail with immutable records"],"categories":["automation-workflow","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":38,"verified":false,"data_access_risk":"high","permissions":["On-premise server infrastructure or private cloud account (AWS PrivateLink, Azure ExpressRoute, or equivalent)","GPU compute (NVIDIA A100 or equivalent) for sub-second inference on models >7B parameters","Data connectors for source systems (Slack API, email server access, database credentials)","Network isolation from public internet or VPN tunnel to WorkHub infrastructure","API credentials or database access for each source system","Sample documents (5-10 per schema variant) for the system to learn extraction patterns","Defined target schema (JSON schema or equivalent) for normalized output","Ongoing manual review of extraction results for quality assurance (recommended 5-10% sampling)","Local LLM model with sufficient context window (2k+ tokens for reasonable document length)","Document content in text format (PDFs require OCR preprocessing)"],"failure_modes":["Local LLM inference typically 2-5x slower than cloud APIs (GPT-4) due to hardware constraints; inference latency scales with model size","Requires dedicated compute infrastructure (GPU/TPU) for reasonable performance; no serverless option for variable workloads","Knowledge consolidation from unstructured sources requires manual schema definition; no automatic format detection across heterogeneous data types","Privacy guarantees only as strong as underlying infrastructure security; misconfiguration of network isolation can expose data","Schema inference accuracy depends on sample quality; ambiguous or sparse examples lead to incorrect field mappings requiring manual correction","Extraction latency scales with document complexity; dense PDFs or images require OCR preprocessing adding 500ms-2s per document","No built-in conflict resolution for duplicate or contradictory data across sources; requires manual deduplication rules","Extraction rules are not portable across schema versions; schema changes require retraining on new samples","Summarization quality degrades for documents >10k tokens; context window limitations force truncation or hierarchical summarization","Extracted insights are only as good as the LLM's understanding; domain-specific insights may be missed or misinterpreted","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.2833333333333333,"quality":0.6799999999999999,"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":"2026-05-24T12:16:34.117Z","last_scraped_at":"2026-04-05T13:23:42.562Z","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=workhub","compare_url":"https://unfragile.ai/compare?artifact=workhub"}},"signature":"Ph7J2diwamHp9miJtoW2AR42m5AR3+Pqo8FCzOp3xs+WdP/YrfPxslvxvW90jmF3wR1cFIx/1s/ticrgFJmvDA==","signedAt":"2026-06-22T13:18:40.294Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/workhub","artifact":"https://unfragile.ai/workhub","verify":"https://unfragile.ai/api/v1/verify?slug=workhub","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"}}