{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"awesome-awesome-rag-production","slug":"awesome-rag-production","name":"Awesome RAG Production","type":"repo","url":"https://github.com/Yigtwxx/Awesome-RAG-Production","page_url":"https://unfragile.ai/awesome-rag-production","categories":["rag-knowledge","deployment-infra"],"tags":[],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"awesome-awesome-rag-production__cap_0","uri":"capability://memory.knowledge.curated.rag.tool.discovery.and.evaluation","name":"curated-rag-tool-discovery-and-evaluation","description":"Provides a systematically organized, community-maintained catalog of production-ready RAG tools, frameworks, and libraries with categorization by function (embedding models, vector databases, retrieval strategies, LLM providers, orchestration frameworks). The curation model relies on GitHub stars, community adoption signals, and maintainer activity to surface tools with proven production viability, enabling builders to quickly identify and compare solutions rather than evaluating from scratch.","intents":["I need to find a vector database that scales to billions of embeddings and integrates with my existing data pipeline","I want to compare embedding model options (open-source vs proprietary) for my specific domain and latency requirements","I'm building a RAG system and need to understand the full ecosystem of tools available before architecting my stack","I need to evaluate which orchestration framework (LangChain, LlamaIndex, etc.) fits my team's Python/TypeScript preference"],"best_for":["ML engineers and architects designing RAG systems from scratch","teams evaluating tool migrations or stack replacements","startups prototyping RAG MVPs with limited evaluation bandwidth"],"limitations":["Curation is manual and asynchronous — may lag behind new tool releases by weeks or months","No automated benchmarking or performance comparison data — relies on external sources","Categorization is static and doesn't capture nuanced trade-offs (e.g., latency vs cost vs accuracy)","No integration testing across tools — compatibility issues between components must be discovered independently"],"requires":["GitHub account or web browser to access the repository","Basic familiarity with RAG architecture concepts (embeddings, vector stores, retrievers)","Ability to evaluate tool maturity by reading READMEs and GitHub metrics"],"input_types":["user search queries or browsing through categorized lists"],"output_types":["structured list of tools with links, descriptions, and metadata","comparison matrices across tool categories"],"categories":["memory-knowledge","search-retrieval"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_1","uri":"capability://planning.reasoning.rag.architecture.pattern.reference","name":"rag-architecture-pattern-reference","description":"Aggregates documented architectural patterns, design decisions, and best practices for building production RAG systems, including chunking strategies, retrieval augmentation approaches (dense vs sparse, hybrid), reranking pipelines, and evaluation frameworks. Serves as a living reference guide that captures lessons learned from deployed systems, enabling builders to avoid common pitfalls and adopt proven patterns without reinventing solutions.","intents":["I need to understand trade-offs between different chunking strategies (fixed-size, semantic, hierarchical) for my document corpus","I want to implement a hybrid retrieval approach combining BM25 and dense embeddings — what's the production-tested pattern?","How do I design a reranking pipeline to improve retrieval quality without adding unacceptable latency?","What evaluation metrics and benchmarks should I use to measure RAG system quality in production?"],"best_for":["ML engineers implementing RAG systems for the first time","teams optimizing existing RAG deployments for quality or latency","architects designing multi-stage retrieval pipelines"],"limitations":["Patterns are descriptive, not prescriptive — no automated tool to apply them to your specific codebase","No domain-specific guidance (e.g., legal documents vs medical records vs code repositories require different strategies)","Patterns may reflect best practices from 6-12 months ago; rapidly evolving field means some recommendations may be superseded","No quantitative benchmarks comparing pattern effectiveness across datasets"],"requires":["Understanding of embedding models and vector similarity","Familiarity with information retrieval concepts (precision, recall, MRR)","Ability to read and interpret technical documentation and research papers"],"input_types":["text descriptions of RAG architecture decisions and trade-offs"],"output_types":["documented patterns with pseudocode or implementation examples","decision trees for selecting between architectural approaches","evaluation frameworks and metrics"],"categories":["planning-reasoning","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_10","uri":"capability://code.generation.editing.rag.fine.tuning.and.domain.adaptation.strategies","name":"rag-fine-tuning-and-domain-adaptation-strategies","description":"Catalogs approaches for adapting RAG systems to specific domains through fine-tuning embedding models, rerankers, and LLMs, as well as techniques for improving retrieval and generation quality for domain-specific use cases. Includes guidance on collecting domain-specific training data, evaluating fine-tuned models, and managing the trade-offs between generic and domain-specific components.","intents":["Should I fine-tune my embedding model for my specific domain, or use a pre-trained model?","How do I collect and prepare training data for fine-tuning domain-specific components?","What's the ROI of fine-tuning a reranker vs using a generic reranker?","How do I evaluate whether fine-tuning improves my RAG system's quality?"],"best_for":["teams building domain-specific RAG systems (legal, medical, financial, etc.)","ML engineers optimizing RAG quality through fine-tuning","organizations with sufficient domain-specific data to support fine-tuning"],"limitations":["Fine-tuning requires significant domain-specific training data — not viable for all use cases","Fine-tuning introduces operational complexity — managing multiple model versions and rollouts","Improvements from fine-tuning are often incremental — may not justify the effort for some use cases","Fine-tuned models may not generalize well to out-of-domain queries or new document types"],"requires":["Domain-specific training data (queries, relevant documents, relevance judgments)","ML infrastructure for fine-tuning (GPUs, training frameworks)","Evaluation methodology to measure fine-tuning impact","Model management and versioning infrastructure"],"input_types":["domain-specific training data and relevance judgments","baseline model performance metrics","domain characteristics and requirements"],"output_types":["fine-tuned embedding models and rerankers","evaluation results comparing fine-tuned vs baseline models","ROI analysis and recommendations","deployment and versioning strategies"],"categories":["code-generation-editing","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_11","uri":"capability://safety.moderation.rag.security.privacy.and.compliance.patterns","name":"rag-security-privacy-and-compliance-patterns","description":"Provides guidance on security, privacy, and compliance considerations for production RAG systems, including data access control, PII handling, audit logging, and regulatory compliance (GDPR, HIPAA, etc.). Addresses unique security challenges in RAG systems such as preventing information leakage through retrieved context and managing sensitive data in vector databases.","intents":["How do I prevent my RAG system from leaking sensitive information through retrieved context?","What access controls should I implement for documents and queries in my RAG system?","How do I handle PII (personally identifiable information) in my RAG pipeline?","What compliance requirements apply to my RAG system (GDPR, HIPAA, SOC 2)?"],"best_for":["security and compliance teams implementing RAG systems","teams building RAG systems for regulated industries (healthcare, finance, legal)","organizations handling sensitive data in RAG pipelines"],"limitations":["Security and compliance requirements are highly domain and jurisdiction-specific — no universal solution","Implementing strong access controls and encryption adds operational complexity and latency","Audit logging and compliance monitoring require significant infrastructure investment","Privacy-preserving techniques (differential privacy, federated learning) may degrade RAG quality"],"requires":["Security and compliance expertise","Understanding of data protection regulations (GDPR, HIPAA, etc.)","Infrastructure for access control, encryption, and audit logging","Privacy-preserving techniques and tools"],"input_types":["system architecture and data flows","regulatory requirements and compliance frameworks","security threat models and risk assessments"],"output_types":["security and privacy architecture diagrams","access control policies and implementations","audit logging and compliance monitoring procedures","incident response and breach notification procedures"],"categories":["safety-moderation","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_2","uri":"capability://data.processing.analysis.rag.evaluation.framework.catalog","name":"rag-evaluation-framework-catalog","description":"Indexes evaluation tools, metrics, and benchmarks for assessing RAG system quality across multiple dimensions (retrieval quality, generation quality, latency, cost). Includes pointers to established benchmarks (TREC, BEIR, custom domain-specific datasets) and evaluation libraries (RAGAS, DeepEval, etc.) that enable builders to measure system performance against production requirements rather than relying on subjective assessment.","intents":["I need to measure whether my retrieval system is finding the right documents — what metrics should I track?","How do I evaluate the quality of generated answers without manual annotation of every query?","I want to set up continuous evaluation in my RAG pipeline to catch quality regressions before they reach users","What benchmarks exist for my specific domain (legal, medical, financial) to validate my system?"],"best_for":["ML engineers implementing observability and quality gates in RAG systems","teams establishing SLOs and performance baselines for RAG deployments","researchers comparing RAG approaches on standardized benchmarks"],"limitations":["Evaluation metrics are often task-specific — no single metric works across all RAG use cases","Automated evaluation (using LLMs to judge answer quality) is itself imperfect and may not correlate with human judgment","Benchmarks may not reflect your specific domain or document distribution — transfer learning from public benchmarks is unreliable","Evaluation infrastructure requires significant engineering effort to integrate into production pipelines"],"requires":["Labeled evaluation datasets (ground truth queries and relevant documents)","Python environment with evaluation libraries (RAGAS, DeepEval, etc.)","Understanding of information retrieval metrics (NDCG, MRR, MAP)","LLM API access for automated evaluation (OpenAI, Anthropic, or local models)"],"input_types":["retrieval results (ranked lists of documents)","generated answers from RAG system","ground truth labels or reference answers"],"output_types":["quantitative metrics (NDCG, MRR, BLEU, ROUGE, etc.)","evaluation reports with quality breakdowns","benchmark comparison matrices"],"categories":["data-processing-analysis","memory-knowledge"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_3","uri":"capability://memory.knowledge.vector.database.and.embedding.model.selection.guide","name":"vector-database-and-embedding-model-selection-guide","description":"Provides comparative information on vector databases (Pinecone, Weaviate, Milvus, Qdrant, etc.) and embedding models (OpenAI, Cohere, open-source options) with guidance on selection criteria including scalability, latency, cost, and integration patterns. Helps builders match their requirements (query throughput, embedding dimension, metadata filtering) to appropriate solutions rather than defaulting to popular choices.","intents":["I need a vector database that supports real-time updates and complex metadata filtering — which options should I evaluate?","Should I use a proprietary embedding model (OpenAI) or fine-tune an open-source model for my domain?","What's the cost difference between managed vector databases (Pinecone) vs self-hosted options (Milvus, Qdrant)?","How do I choose between vector databases based on query latency requirements and scale?"],"best_for":["architects selecting core infrastructure for RAG systems","teams evaluating cost-performance trade-offs for vector storage","engineers migrating between vector database providers"],"limitations":["Comparative data is static and doesn't reflect real-time performance changes or new releases","Benchmarks are often vendor-provided and may not be independent or reproducible","No guidance on operational complexity (backup, disaster recovery, monitoring) which varies significantly across options","Embedding model selection depends heavily on domain-specific factors not captured in generic comparisons"],"requires":["Understanding of vector similarity search and approximate nearest neighbor algorithms","Knowledge of your system's query throughput and latency requirements","Familiarity with embedding dimensions and metadata filtering needs"],"input_types":["system requirements (scale, latency, cost budget)","domain characteristics (document types, query patterns)"],"output_types":["comparison matrices of vector databases and embedding models","selection decision trees based on requirements","cost and performance estimates"],"categories":["memory-knowledge","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_4","uri":"capability://automation.workflow.rag.deployment.and.scaling.patterns","name":"rag-deployment-and-scaling-patterns","description":"Catalogs deployment architectures, scaling strategies, and operational patterns for production RAG systems, including containerization approaches, load balancing for retrieval, caching strategies, and multi-region deployment. Enables builders to move from prototype to production by providing reference architectures that address operational concerns like availability, cost optimization, and monitoring.","intents":["How do I scale my RAG system to handle 1000s of concurrent queries without overwhelming the vector database?","What caching strategies reduce embedding computation and vector database queries in production?","How do I deploy a RAG system across multiple regions for low-latency access?","What monitoring and alerting should I set up for a production RAG pipeline?"],"best_for":["DevOps engineers and platform teams deploying RAG systems","teams scaling RAG systems from prototype to production","architects designing multi-region or high-availability RAG deployments"],"limitations":["Deployment patterns are infrastructure-specific (Kubernetes, serverless, traditional VMs) — no one-size-fits-all solution","Scaling bottlenecks vary by system design (retrieval-bound vs generation-bound vs embedding-bound) — patterns must be customized","Cost optimization trade-offs are highly dependent on query patterns and SLOs — generic guidance may not apply","Monitoring and observability requirements evolve as systems mature — patterns may need continuous refinement"],"requires":["Containerization knowledge (Docker, Kubernetes or equivalent)","Understanding of distributed systems concepts (load balancing, caching, replication)","Monitoring and observability tools (Prometheus, Datadog, etc.)","Infrastructure as code experience (Terraform, CloudFormation, etc.)"],"input_types":["system requirements (throughput, latency, availability SLOs)","infrastructure constraints (budget, regions, compliance)"],"output_types":["reference architectures with deployment diagrams","scaling strategies with cost-performance trade-offs","monitoring and alerting configurations","disaster recovery and failover patterns"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_5","uri":"capability://tool.use.integration.rag.framework.and.orchestration.tool.comparison","name":"rag-framework-and-orchestration-tool-comparison","description":"Provides comparative analysis of RAG orchestration frameworks (LangChain, LlamaIndex, Haystack, etc.) with guidance on framework selection based on use case, language preference, and integration needs. Captures architectural differences in how frameworks handle retrieval, generation, and state management, enabling builders to select frameworks that match their development velocity and operational requirements.","intents":["Should I use LangChain or LlamaIndex for my RAG system — what are the architectural differences?","I need a framework that supports complex multi-step retrieval pipelines with custom logic — which options are most flexible?","What's the learning curve and community support for different RAG frameworks?","How do different frameworks handle state management and memory in production?"],"best_for":["developers building RAG applications and selecting foundational frameworks","teams evaluating framework migrations or replacements","architects designing RAG systems with specific integration requirements"],"limitations":["Framework landscapes evolve rapidly — comparisons become outdated quickly as new versions are released","Framework selection is often path-dependent — switching frameworks mid-project is costly","Abstraction levels vary significantly — some frameworks hide complexity while others expose it, affecting both ease-of-use and control","Community size and ecosystem maturity vary, affecting availability of integrations and third-party tools"],"requires":["Proficiency in Python or TypeScript (depending on framework choice)","Understanding of RAG architecture concepts","Familiarity with LLM APIs and integration patterns"],"input_types":["system requirements (language, integration needs, complexity)","team preferences (development velocity vs control)"],"output_types":["framework comparison matrices","architecture diagrams showing framework design patterns","code examples demonstrating framework usage","decision trees for framework selection"],"categories":["tool-use-integration","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_6","uri":"capability://data.processing.analysis.rag.cost.optimization.and.economics.guide","name":"rag-cost-optimization-and-economics-guide","description":"Aggregates strategies and tools for optimizing RAG system costs across embedding computation, vector database storage and queries, and LLM inference. Includes cost modeling approaches, trade-off analysis between proprietary and open-source components, and techniques for reducing operational expenses without sacrificing quality (caching, batching, quantization).","intents":["How can I reduce embedding computation costs — should I cache embeddings or use cheaper embedding models?","What's the cost impact of different vector database choices at scale (1M+ documents)?","How do I optimize LLM inference costs in my RAG pipeline without degrading answer quality?","What's the total cost of ownership for my RAG system including infrastructure, APIs, and operational overhead?"],"best_for":["startups and teams with cost-sensitive RAG deployments","finance and operations teams optimizing RAG system budgets","engineers implementing cost monitoring and optimization in production"],"limitations":["Cost models are highly dependent on usage patterns (query volume, document size, update frequency) — generic estimates may be inaccurate","Pricing changes frequently across vendors — cost comparisons become stale quickly","Cost-quality trade-offs are domain-specific — optimizations that work for one use case may not apply to others","Hidden costs (operational overhead, monitoring, compliance) are often underestimated in initial cost models"],"requires":["Understanding of your system's resource consumption (API calls, storage, compute)","Pricing information from vendors (embedding models, vector databases, LLM providers)","Ability to measure and track actual costs in production"],"input_types":["system architecture and resource usage patterns","vendor pricing and rate structures","quality requirements and acceptable trade-offs"],"output_types":["cost models and projections","cost-quality trade-off matrices","optimization recommendations with estimated savings","cost monitoring dashboards and alerts"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_7","uri":"capability://data.processing.analysis.rag.data.pipeline.and.ingestion.patterns","name":"rag-data-pipeline-and-ingestion-patterns","description":"Catalogs data ingestion, preprocessing, and pipeline patterns for RAG systems, including document parsing, chunking strategies, metadata extraction, and incremental updates. Provides guidance on building robust data pipelines that handle diverse document formats, maintain data quality, and support continuous indexing without system downtime.","intents":["How do I build a data pipeline that ingests documents from multiple sources (PDFs, web pages, databases) into my RAG system?","What chunking strategy should I use for different document types (code, legal documents, scientific papers)?","How do I extract and maintain metadata (source, date, author) through the ingestion pipeline?","How do I update my vector index with new documents without rebuilding from scratch?"],"best_for":["data engineers building data pipelines for RAG systems","teams managing large document corpora with frequent updates","architects designing end-to-end RAG systems including data infrastructure"],"limitations":["Document parsing is format-specific — no universal solution handles all document types equally well","Chunking strategies are domain-dependent — optimal chunk size and strategy varies by document type and retrieval use case","Data quality issues (duplicates, corrupted documents, metadata errors) require domain-specific handling","Incremental indexing adds complexity — managing deltas and updates requires careful state management"],"requires":["Data engineering experience with ETL/ELT pipelines","Document parsing libraries (PyPDF2, pdfplumber, BeautifulSoup, etc.)","Understanding of chunking trade-offs (size, overlap, semantic boundaries)","Orchestration tools (Airflow, Prefect, etc.) for pipeline management"],"input_types":["raw documents in various formats (PDF, HTML, Markdown, JSON, etc.)","metadata and source information","update schedules and incremental change feeds"],"output_types":["chunked documents with metadata","embeddings ready for vector database ingestion","data quality metrics and validation reports","pipeline logs and monitoring data"],"categories":["data-processing-analysis","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_8","uri":"capability://text.generation.language.rag.context.window.and.prompt.engineering.guide","name":"rag-context-window-and-prompt-engineering-guide","description":"Provides strategies for effective prompt engineering in RAG systems, including context window management, prompt templates, and techniques for improving generation quality given retrieved context. Covers trade-offs between context length and cost, strategies for handling irrelevant or conflicting retrieved documents, and methods for guiding LLM behavior within RAG pipelines.","intents":["How do I structure prompts to effectively use retrieved context without overwhelming the LLM?","What's the optimal context window size for my RAG system given cost and quality constraints?","How do I handle cases where retrieved documents are irrelevant or contradictory?","What prompt engineering techniques improve answer quality in RAG systems?"],"best_for":["ML engineers and prompt engineers optimizing RAG generation quality","teams fine-tuning RAG systems for specific domains or use cases","developers building RAG applications with quality requirements"],"limitations":["Prompt engineering is largely empirical — techniques that work for one domain may not transfer to others","LLM behavior varies across models and versions — prompts require retuning when switching models","Context window management is a cost-quality trade-off — optimal window size depends on specific use case and cost constraints","Handling irrelevant or conflicting context is challenging — no universal solution works across all scenarios"],"requires":["Access to LLM APIs or local models for experimentation","Understanding of prompt engineering principles and techniques","Ability to evaluate generation quality (manual review, automated metrics, user feedback)"],"input_types":["retrieved documents and context","user queries","domain-specific knowledge and constraints"],"output_types":["prompt templates and examples","context formatting strategies","generation quality metrics and evaluation results","cost-quality trade-off analysis"],"categories":["text-generation-language","planning-reasoning"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"awesome-awesome-rag-production__cap_9","uri":"capability://automation.workflow.rag.monitoring.observability.and.debugging.toolkit","name":"rag-monitoring-observability-and-debugging-toolkit","description":"Aggregates monitoring, observability, and debugging tools and patterns for production RAG systems, including metrics for retrieval quality, generation quality, latency, and cost. Provides guidance on setting up alerts, dashboards, and debugging workflows to identify and resolve issues in production RAG pipelines.","intents":["What metrics should I monitor to detect quality degradation in my RAG system?","How do I debug cases where my RAG system returns irrelevant answers?","What observability infrastructure do I need for a production RAG system?","How do I set up alerts for SLO violations in my RAG pipeline?"],"best_for":["DevOps and SRE teams operating production RAG systems","ML engineers implementing observability in RAG pipelines","teams establishing SLOs and quality baselines for RAG systems"],"limitations":["Observability requirements are system-specific — no one-size-fits-all monitoring strategy","Debugging RAG systems is complex — issues can originate in retrieval, generation, or data quality","Automated quality detection is imperfect — many issues require manual investigation or user feedback","Observability infrastructure adds operational overhead and cost"],"requires":["Monitoring and observability tools (Prometheus, Datadog, ELK, etc.)","Logging infrastructure for RAG pipeline components","Ability to define and measure SLOs for RAG systems","Understanding of RAG failure modes and debugging techniques"],"input_types":["RAG system logs and traces","performance metrics (latency, throughput, cost)","quality metrics (retrieval accuracy, generation quality)"],"output_types":["monitoring dashboards and alerts","debugging guides and runbooks","SLO definitions and tracking","incident response procedures"],"categories":["automation-workflow","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":26,"verified":false,"data_access_risk":"high","permissions":["GitHub account or web browser to access the repository","Basic familiarity with RAG architecture concepts (embeddings, vector stores, retrievers)","Ability to evaluate tool maturity by reading READMEs and GitHub metrics","Understanding of embedding models and vector similarity","Familiarity with information retrieval concepts (precision, recall, MRR)","Ability to read and interpret technical documentation and research papers","Domain-specific training data (queries, relevant documents, relevance judgments)","ML infrastructure for fine-tuning (GPUs, training frameworks)","Evaluation methodology to measure fine-tuning impact","Model management and versioning infrastructure"],"failure_modes":["Curation is manual and asynchronous — may lag behind new tool releases by weeks or months","No automated benchmarking or performance comparison data — relies on external sources","Categorization is static and doesn't capture nuanced trade-offs (e.g., latency vs cost vs accuracy)","No integration testing across tools — compatibility issues between components must be discovered independently","Patterns are descriptive, not prescriptive — no automated tool to apply them to your specific codebase","No domain-specific guidance (e.g., legal documents vs medical records vs code repositories require different strategies)","Patterns may reflect best practices from 6-12 months ago; rapidly evolving field means some recommendations may be superseded","No quantitative benchmarks comparing pattern effectiveness across datasets","Fine-tuning requires significant domain-specific training data — not viable for all use cases","Fine-tuning introduces operational complexity — managing multiple model versions and rollouts","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.34,"ecosystem":0.49999999999999994,"match_graph":0.25,"freshness":0.6,"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":"2026-06-17T09:51:02.371Z","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=awesome-rag-production","compare_url":"https://unfragile.ai/compare?artifact=awesome-rag-production"}},"signature":"kXbVt/A1rLJzwqqvIx75DCiTkbFnCZlso+jeih/6Ux7Or0G7ydixZ9D/vDmm86xD6NrBkZZqIHeTOlqhWKupBg==","signedAt":"2026-06-20T19:01:46.060Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/awesome-rag-production","artifact":"https://unfragile.ai/awesome-rag-production","verify":"https://unfragile.ai/api/v1/verify?slug=awesome-rag-production","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"}}