{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"awesome-agentset","slug":"agentset","name":"Agentset","type":"repo","url":"https://agentset.ai/","page_url":"https://unfragile.ai/agentset","categories":["frameworks-sdks","rag-knowledge"],"tags":[],"pricing":{"model":"unknown","free":false,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"awesome-agentset__cap_0","uri":"capability://search.retrieval.semantic.search.with.hybrid.reranking","name":"semantic-search-with-hybrid-reranking","description":"Executes vector-based semantic search across ingested documents combined with BM25 keyword matching, then applies a reranking algorithm to surface most relevant results. The system converts user queries to embeddings, searches a vector database (Pinecone or Qdrant), retrieves candidate documents, and reranks them using a learned-to-rank model before returning cited sources. This hybrid approach balances semantic understanding with keyword precision.","intents":["I want to find the most relevant documents from my knowledge base without writing exact keyword queries","I need search results ranked by relevance, not just keyword matching","I want to retrieve documents with source attribution for compliance and verification"],"best_for":["teams building internal knowledge bases or customer support systems","enterprises requiring cited sources for regulatory compliance","developers integrating RAG into LLM applications"],"limitations":["Reranking algorithm specifics not documented — unclear if it uses cross-encoder models or proprietary approach","No control over embedding model selection exposed in public documentation","Latency of hybrid search + reranking not published; likely adds 200-500ms per query","Vector database choice (Pinecone vs Qdrant) affects cost and performance but selection criteria not documented"],"requires":["Ingested documents in supported format (22+ formats including PDF, DOCX, images)","Active Agentset namespace with configured embedding model","API key or SDK authentication"],"input_types":["text query (natural language)","metadata filters (key-value pairs)"],"output_types":["ranked document chunks with relevance scores","source citations (document name, page number, URL)","metadata associated with retrieved chunks"],"categories":["search-retrieval","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_1","uri":"capability://planning.reasoning.multi.hop.document.reasoning","name":"multi-hop-document-reasoning","description":"Enables answering questions that require retrieving and reasoning across multiple documents sequentially. The system performs iterative retrieval: initial query retrieves relevant documents, LLM generates follow-up queries based on retrieved context, system retrieves additional documents, and final answer synthesizes information across all retrieved sources. This is benchmarked on MultiHopQA, indicating support for 2-3 hop reasoning chains.","intents":["I need to answer questions that require information from multiple documents","I want the system to automatically find related documents without me specifying each search","I need to trace the reasoning path showing which documents contributed to the answer"],"best_for":["financial analysis teams answering questions across multiple reports","legal teams researching precedents across case documents","research teams synthesizing findings from multiple papers"],"limitations":["Hop depth not documented — unclear if limited to 2-3 hops or supports deeper chains","No explicit control over reasoning strategy (greedy vs exhaustive search)","Reasoning process is implicit in LLM behavior — not exposed as structured chain-of-thought","Performance degrades with document count; no published latency benchmarks for multi-hop queries","Cannot branch reasoning (e.g., explore multiple hypotheses in parallel)"],"requires":["Minimum 3-5 documents in knowledge base for meaningful multi-hop reasoning","LLM model configured with sufficient context window (4K+ tokens recommended)","Metadata or semantic similarity enabling document linking"],"input_types":["natural language question requiring cross-document reasoning"],"output_types":["synthesized answer with citations to multiple source documents","implicit reasoning chain (not explicitly exposed)"],"categories":["planning-reasoning","search-retrieval"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_10","uri":"capability://automation.workflow.webhook.based.ingestion.event.tracking","name":"webhook-based-ingestion-event-tracking","description":"Provides webhook callbacks for document ingestion lifecycle events (started, completed, failed), enabling external systems to track ingestion status and trigger downstream workflows. The system sends HTTP POST requests to configured webhook URLs with event metadata (document ID, status, error details), allowing asynchronous monitoring without polling the API.","intents":["I want to be notified when documents finish ingesting so I can trigger downstream processes","I need to track ingestion failures and retry failed documents","I want to log ingestion events for audit and compliance purposes"],"best_for":["teams with automated document processing pipelines","enterprises requiring audit trails for document ingestion","applications needing to coordinate ingestion with other systems"],"limitations":["Webhook event types not fully documented — unclear what events are supported beyond ingestion status","Retry logic not documented — unclear if failed webhooks are retried","Webhook authentication not documented — unclear if signed requests or API key validation","Payload schema not documented — unclear what metadata is included in webhook events","No webhook management API documented — unclear how to list, update, or delete webhooks"],"requires":["Enterprise tier (webhooks appear to be enterprise-only feature)","Public HTTPS endpoint to receive webhooks","Webhook URL configuration in Agentset dashboard"],"input_types":["webhook configuration (URL, event types)"],"output_types":["HTTP POST requests with event metadata"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_11","uri":"capability://automation.workflow.bring.your.own.cloud.and.on.premise.deployment","name":"bring-your-own-cloud-and-on-premise-deployment","description":"Enables enterprise customers to deploy Agentset in their own cloud infrastructure (AWS, Azure, GCP) or on-premise data centers, maintaining full data sovereignty and control. The deployment includes all components (API, vector database, LLM integration) and can be configured for high availability and disaster recovery. Data never leaves the customer's infrastructure.","intents":["I need to keep all data within my organization's infrastructure for compliance or security","I want to deploy Agentset in my existing cloud account to avoid vendor lock-in","I need to meet data residency requirements for regulated industries"],"best_for":["enterprises with strict data sovereignty requirements (HIPAA, GDPR, financial services)","organizations with existing cloud infrastructure and DevOps teams","teams requiring custom security configurations or air-gapped deployments"],"limitations":["BYOC and on-premise are enterprise-only features — no pricing or SLA documentation","Deployment architecture not documented — unclear what components are included","Infrastructure requirements not documented — unclear compute, storage, network requirements","Maintenance and update procedures not documented","Support model for self-hosted deployments not documented"],"requires":["Enterprise tier subscription","Cloud infrastructure (AWS, Azure, GCP) or on-premise data center","DevOps/infrastructure team for deployment and maintenance","Network connectivity for LLM provider APIs (if using external LLMs)"],"input_types":["deployment configuration (cloud provider, region, scaling parameters)"],"output_types":["fully deployed Agentset instance in customer infrastructure"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_12","uri":"capability://automation.workflow.per.page.ingestion.pricing.with.unlimited.retrieval","name":"per-page-ingestion-pricing-with-unlimited-retrieval","description":"Uses a consumption-based pricing model where customers pay per document page ingested ($0.01/page on Pro tier after 10,000 included pages) but have unlimited retrieval queries. This decouples ingestion costs from query volume, making the service cost-predictable for high-query-volume use cases. Free tier includes 1,000 pages and 10,000 retrievals/month.","intents":["I want to understand the cost structure before building a RAG application","I need predictable costs for a high-query-volume knowledge base","I want to avoid per-query pricing that scales with user adoption"],"best_for":["teams with large document collections but variable query volume","applications expecting high user adoption and query growth","enterprises with predictable document ingestion but unpredictable usage"],"limitations":["Connector costs ($100/month per connector) add significant overhead for multi-source setups","Free tier limits (1,000 pages, 10,000 retrievals/month) are restrictive for production use","Pro tier pricing ($0.01/page) can be expensive for large document collections (100,000 pages = $1,000)","No volume discounts documented","Cost of external LLM API calls (OpenAI, Anthropic, etc.) not included in Agentset pricing"],"requires":["Understanding of document page count (how PDFs are counted not documented)","Budget for connector costs if using multiple data sources"],"input_types":["pricing tier selection (Free, Pro, Enterprise)"],"output_types":["cost estimates based on page count and connector usage"],"categories":["automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_13","uri":"capability://safety.moderation.compliance.and.security.features.for.enterprise","name":"compliance-and-security-features-for-enterprise","description":"Provides enterprise-grade security and compliance features including SOC 2 certification, HIPAA compliance, GDPR data handling, and audit logging. The platform supports role-based access control, data encryption at rest and in transit, and compliance reporting. Specific implementation details are not publicly documented but are available under NDA for enterprise customers.","intents":["I need to ensure my knowledge base meets regulatory compliance requirements","I want audit trails and access logs for compliance reporting","I need to restrict access to sensitive documents by user role"],"best_for":["healthcare organizations handling PHI (HIPAA)","financial services firms with regulatory requirements","enterprises in regulated industries (legal, government)"],"limitations":["Compliance features are enterprise-only — no public documentation of specific controls","Audit logging scope not documented — unclear what events are logged","RBAC implementation not documented — unclear granularity of access control","Data retention and deletion policies not documented","Compliance certifications (SOC 2, HIPAA) not publicly verified"],"requires":["Enterprise tier subscription","Compliance requirements documentation for vendor assessment","NDA for detailed security documentation"],"input_types":["compliance requirements and audit requests"],"output_types":["compliance reports, audit logs, security documentation"],"categories":["safety-moderation","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_2","uri":"capability://data.processing.analysis.multimodal.document.ingestion.and.retrieval","name":"multimodal-document-ingestion-and-retrieval","description":"Processes 22+ file formats including PDFs, images (PNG, JPEG), tables (XLSX), presentations (PPTX), and structured data (CSV, XML, JSON) into a unified searchable index. The system extracts text from images using OCR, parses table structures, preserves formatting metadata, and creates embeddings for both text and visual content. Retrieved results include the original visual elements alongside text, enabling questions about charts, diagrams, and images.","intents":["I need to search across documents containing images, charts, and tables, not just text","I want to ask questions about visual content in PDFs and presentations","I need to ingest mixed-format data (some PDFs, some spreadsheets, some images) into a single searchable knowledge base"],"best_for":["teams managing technical documentation with diagrams and screenshots","financial/legal teams processing reports with tables and charts","enterprises with heterogeneous document formats (legacy systems, multiple departments)"],"limitations":["OCR quality not documented — unclear accuracy on handwritten text or low-resolution images","Table extraction may fail on complex nested tables or merged cells","Video and audio not supported — only static images","Image retrieval relies on OCR + text embedding, not visual embeddings (CLIP-style), limiting image-to-image search","Custom format support (beyond 22 listed) requires enterprise plan"],"requires":["File size limits not documented (typical SaaS: 10-100MB per file)","Supported format (PDF, DOCX, XLSX, PPTX, PNG, JPEG, CSV, HTML, MD, TXT, XML, JSON, etc.)","Pro tier: $0.01 per page ingested (beyond 10,000 included pages)"],"input_types":["file upload (22+ formats)","URL crawling (HTML, PDF)","connector integration (Google Drive, SharePoint, Notion)"],"output_types":["extracted text with source attribution","visual elements (images, tables) with bounding boxes or references","structured metadata (page number, section, format type)"],"categories":["data-processing-analysis","image-visual"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_3","uri":"capability://search.retrieval.metadata.filtering.and.faceted.search","name":"metadata-filtering-and-faceted-search","description":"Enables filtering retrieved documents by custom metadata (key-value pairs) attached during ingestion, allowing queries like 'find documents from Q3 2024 with department=finance'. Metadata is indexed alongside embeddings, enabling combined semantic + metadata filtering in a single query. Supports boolean operators (AND, OR, NOT) and range queries on numeric metadata.","intents":["I want to search only within documents from a specific time period or department","I need to exclude certain document types or sources from search results","I want to combine semantic search with structured filtering (e.g., 'find recent documents about X')"],"best_for":["multi-tenant SaaS platforms isolating data by customer or workspace","enterprises with document versioning or temporal queries","teams managing documents across multiple projects or departments"],"limitations":["Metadata schema not enforced — no validation of metadata types or required fields","Query syntax for complex filters not documented (unclear if supports nested boolean logic)","No aggregation/faceting API documented (e.g., 'count documents by department')","Metadata cardinality limits not specified — unclear performance with high-cardinality fields","No metadata update API documented — metadata appears immutable after ingestion"],"requires":["Metadata attached during document ingestion via `config.metadata` parameter","Metadata keys must be strings; value types not documented"],"input_types":["metadata key-value pairs (during ingestion)","filter expressions (during search)"],"output_types":["filtered document chunks matching both semantic and metadata criteria","metadata values included in search results"],"categories":["search-retrieval","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_4","uri":"capability://text.generation.language.conversational.rag.with.context.management","name":"conversational-rag-with-context-management","description":"Maintains multi-turn conversation state where each user message is augmented with retrieved context from the knowledge base before being sent to the LLM. The system retrieves relevant documents for each turn, appends them to the conversation history, and passes the enriched context to the LLM for response generation. This enables coherent multi-turn Q&A where the LLM can reference both previous conversation turns and retrieved documents.","intents":["I want to build a chatbot that answers follow-up questions about documents","I need the system to remember previous questions in a conversation while retrieving new documents","I want to have a natural back-and-forth conversation with a document-aware assistant"],"best_for":["customer support chatbots answering questions from knowledge bases","internal documentation assistants for employees","research assistants helping users explore document collections"],"limitations":["Context window management not documented — unclear how conversation history is truncated for long conversations","No explicit conversation persistence API documented — unclear if conversations are stored or ephemeral","Conversation isolation not documented — unclear if multi-user conversations are supported","No conversation branching or alternative response exploration","Token usage per turn not tracked or limited in documentation"],"requires":["LLM model with sufficient context window (4K+ tokens for typical conversations)","Conversation session identifier (implementation details not documented)"],"input_types":["user message (text)","conversation history (implicit, managed by system)"],"output_types":["assistant response with citations to retrieved documents","conversation turn with metadata (timestamp, retrieved sources)"],"categories":["text-generation-language","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_5","uri":"capability://automation.workflow.connector.based.continuous.document.sync","name":"connector-based-continuous-document-sync","description":"Integrates with external data sources (Google Drive, SharePoint, Notion) via pre-configured connectors that automatically crawl and ingest documents on a schedule. The system maintains a mapping between source documents and ingested chunks, enabling automatic updates when source documents change. Connectors handle authentication, pagination, and format conversion without requiring manual intervention.","intents":["I want my knowledge base to automatically stay in sync with documents in Google Drive or SharePoint","I need to ingest documents from multiple sources without manual uploads","I want to avoid re-uploading documents every time they're updated"],"best_for":["enterprises with centralized document repositories (SharePoint, Google Workspace)","teams using Notion as a knowledge base that want to make it searchable","organizations needing continuous document synchronization"],"limitations":["Connector cost: $100/month per connector on Pro tier (significant overhead for multi-source setups)","Supported sources limited to Google Drive, SharePoint, Notion — no Slack, GitHub, Jira, or custom APIs","Sync frequency not documented — unclear if real-time or batch (hourly, daily)","Deletion handling not documented — unclear if deleted source documents are removed from index","No connector monitoring or error alerting documented"],"requires":["Pro tier or higher ($100/connector/month)","OAuth credentials for source platform (Google, Microsoft, Notion)","Source documents in supported format"],"input_types":["connector configuration (source URL, credentials, folder path)"],"output_types":["automatically ingested documents with source tracking","sync status and logs (implementation details not documented)"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_6","uri":"capability://tool.use.integration.model.agnostic.llm.integration","name":"model-agnostic-llm-integration","description":"Abstracts LLM provider selection, allowing users to configure different LLM backends (OpenAI, Anthropic Claude, Google AI, xAI Grok, Azure, Cohere, Qwen, Mistral, DeepSeek) without changing application code. The system handles provider-specific API differences, token counting, and response formatting transparently. Users specify model via configuration, and the platform routes requests to the appropriate provider.","intents":["I want to switch between LLM providers without rewriting my application","I need to use a specific LLM (e.g., Claude for better reasoning) without vendor lock-in","I want to compare outputs from different models on the same knowledge base"],"best_for":["teams evaluating multiple LLM providers","enterprises with multi-cloud or multi-vendor strategies","developers building LLM applications that need provider flexibility"],"limitations":["Model defaults not documented — unclear which model is used if none specified","Provider-specific features (e.g., vision, function calling) not abstracted — may require conditional code","Token counting and cost tracking not unified across providers","No automatic fallback if primary provider is unavailable","Embedding model selection not exposed in documentation — unclear if it's configurable per provider"],"requires":["API key for selected LLM provider (OpenAI, Anthropic, Google, etc.)","Model name/identifier (e.g., 'gpt-4', 'claude-3-opus')"],"input_types":["model configuration (provider, model ID, API key)"],"output_types":["LLM response with provider-agnostic formatting"],"categories":["tool-use-integration","text-generation-language"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_7","uri":"capability://tool.use.integration.typescript.and.python.sdk.with.ai.sdk.integration","name":"typescript-and-python-sdk-with-ai-sdk-integration","description":"Provides TypeScript and Python SDKs with native bindings to Vercel's AI SDK, enabling seamless integration into existing AI applications. The SDK abstracts HTTP calls to the Agentset API, handles authentication, manages request/response serialization, and provides type-safe interfaces (TypeScript). AI SDK integration enables use of Agentset as a tool within AI SDK agent frameworks.","intents":["I want to integrate Agentset into my Node.js or Python application without writing HTTP clients","I need type safety and IDE autocomplete for Agentset API calls","I want to use Agentset as a tool within Vercel's AI SDK agent framework"],"best_for":["Node.js/TypeScript developers building LLM applications","Python developers integrating RAG into existing applications","teams using Vercel's AI SDK for agent development"],"limitations":["Python SDK feature parity with TypeScript SDK not documented","No async/await patterns documented for Python SDK","AI SDK integration limited to Vercel's framework — no LangChain or LlamaIndex adapters documented","SDK version management and deprecation policy not documented","No offline mode or local caching in SDK"],"requires":["Node.js 18+ (TypeScript SDK) or Python 3.9+ (Python SDK)","API key for Agentset authentication","Vercel AI SDK 3.0+ (for AI SDK integration)"],"input_types":["SDK method calls with typed parameters"],"output_types":["typed response objects with full IDE autocomplete"],"categories":["tool-use-integration","code-generation-editing"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_8","uri":"capability://tool.use.integration.model.context.protocol.server.for.external.app.integration","name":"model-context-protocol-server-for-external-app-integration","description":"Exposes Agentset as an MCP (Model Context Protocol) server, enabling external applications and LLM clients to query the knowledge base through a standardized protocol. The MCP server implements Agentset's search and retrieval capabilities as MCP tools, allowing any MCP-compatible client (Claude, other LLMs, custom agents) to access the knowledge base without direct API integration.","intents":["I want to use Agentset as a knowledge source within Claude or other MCP-compatible LLMs","I need to expose my knowledge base to external tools without building custom integrations","I want to enable other teams' applications to query my knowledge base via a standard protocol"],"best_for":["teams using Claude with MCP support","enterprises standardizing on MCP for tool integration","organizations sharing knowledge bases across multiple applications"],"limitations":["MCP server capabilities not fully documented — unclear which Agentset features are exposed as MCP tools","MCP version not specified — unclear if MCP 1.0 or newer","Client support limited to MCP-compatible applications — not all LLMs support MCP yet","No authentication/authorization model documented for MCP access","Rate limiting and quota enforcement not documented for MCP clients"],"requires":["MCP-compatible client (Claude, custom agent, etc.)","Agentset namespace with configured knowledge base","MCP server running (deployment details not documented)"],"input_types":["MCP tool calls from client applications"],"output_types":["MCP-formatted tool results with retrieved documents"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-agentset__cap_9","uri":"capability://planning.reasoning.enterprise.deep.research.mode","name":"enterprise-deep-research-mode","description":"An enterprise-tier feature enabling extended multi-step reasoning over documents with configurable depth and breadth. The system performs iterative retrieval and synthesis with explicit reasoning steps, potentially including hypothesis generation, evidence gathering, and conclusion refinement. Specific implementation details are not publicly documented, but benchmarking on FinanceBench suggests capability for complex financial analysis.","intents":["I need to perform deep analysis across many documents with explicit reasoning steps","I want the system to explore multiple hypotheses and synthesize findings","I need to generate research reports with detailed reasoning chains"],"best_for":["financial analysts performing multi-document research","legal teams conducting discovery across case documents","research teams synthesizing findings from large document collections"],"limitations":["Feature is enterprise-only — no public documentation of capabilities or pricing","Reasoning depth and breadth not configurable in public API","Latency not documented — likely significantly higher than standard retrieval","Cost model not documented — may have per-query or per-token charges","No control over reasoning strategy or intermediate steps exposed"],"requires":["Enterprise tier subscription","Large knowledge base (100+ documents recommended)","Complex questions requiring multi-step reasoning"],"input_types":["natural language research question"],"output_types":["detailed research synthesis with reasoning steps (format not documented)"],"categories":["planning-reasoning","search-retrieval"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":28,"verified":false,"data_access_risk":"high","permissions":["Ingested documents in supported format (22+ formats including PDF, DOCX, images)","Active Agentset namespace with configured embedding model","API key or SDK authentication","Minimum 3-5 documents in knowledge base for meaningful multi-hop reasoning","LLM model configured with sufficient context window (4K+ tokens recommended)","Metadata or semantic similarity enabling document linking","Enterprise tier (webhooks appear to be enterprise-only feature)","Public HTTPS endpoint to receive webhooks","Webhook URL configuration in Agentset dashboard","Enterprise tier subscription"],"failure_modes":["Reranking algorithm specifics not documented — unclear if it uses cross-encoder models or proprietary approach","No control over embedding model selection exposed in public documentation","Latency of hybrid search + reranking not published; likely adds 200-500ms per query","Vector database choice (Pinecone vs Qdrant) affects cost and performance but selection criteria not documented","Hop depth not documented — unclear if limited to 2-3 hops or supports deeper chains","No explicit control over reasoning strategy (greedy vs exhaustive search)","Reasoning process is implicit in LLM behavior — not exposed as structured chain-of-thought","Performance degrades with document count; no published latency benchmarks for multi-hop queries","Cannot branch reasoning (e.g., explore multiple hypotheses in parallel)","Webhook event types not fully documented — unclear what events are supported beyond ingestion status","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.5,"ecosystem":0.35000000000000003,"match_graph":0.25,"freshness":0.75,"weights":{"adoption":0.3,"quality":0.2,"ecosystem":0.15,"match_graph":0.3,"freshness":0.05}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-06-17T09:51:02.370Z","last_scraped_at":"2026-05-03T14:00:20.516Z","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=agentset","compare_url":"https://unfragile.ai/compare?artifact=agentset"}},"signature":"obpbb4YxoCrCtBLFESNDqPbijk/PpWFzTjrDQ8m1q7+4fumCud5YcM8LNJrkxlBt64cwwCaasXxz7uqrJjCmBw==","signedAt":"2026-06-21T01:36:55.522Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/agentset","artifact":"https://unfragile.ai/agentset","verify":"https://unfragile.ai/api/v1/verify?slug=agentset","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"}}