[{"data":1,"prerenderedAt":824},["ShallowReactive",2],{"blog-en-cncf-graduated-projects-2025-overview":3,"blog-related-en-cncf-graduated-projects-2025-overview":771,"blog-en-cncf-graduated-projects-2025-overview-alt":757},{"id":4,"title":5,"author":6,"body":7,"date":752,"description":753,"extension":754,"image":708,"locale":755,"meta":756,"navigation":757,"path":758,"seo":759,"stem":760,"tags":761,"__hash__":770},"blog\u002Fblog\u002Fen\u002Fcncf-graduated-projects-2025-overview.md","CNCF Graduated Projects 2025: Production-Ready Open Source You Can Trust","Kubo Team",{"type":8,"value":9,"toc":707},"minimark",[10,33,38,47,81,90,98,102,107,115,119,133,137,145,149,163,167,175,179,187,193,197,201,214,218,226,230,238,242,250,256,260,264,272,276,284,288,296,300,308,312,320,324,328,336,340,348,352,360,364,372,376,390,394,407,411,415,429,433,447,451,459,463,471,475,483,487,529,533,539,629,638,645,649,652,685,695],[11,12,13,14,18,19,26,27,32],"p",{},"When adopting cloud-native technologies, one of the most critical questions is \"which open source projects can be trusted in production?\" The ",[15,16,17],"strong",{},"Graduated"," status from ",[20,21,25],"a",{"href":22,"rel":23},"https:\u002F\u002Fwww.cncf.io\u002F",[24],"nofollow","CNCF (Cloud Native Computing Foundation)"," is the highest level of certification, proving that a project is mature and has extensive production adoption. ",[20,28,31],{"href":29,"rel":30},"https:\u002F\u002Fkubo.hexabase.io\u002F",[24],"Kubo",", as a K3s-based Kubernetes platform, has strong affinity with these CNCF Graduated projects. This article provides a comprehensive category-by-category guide to all Graduated projects as of 2025.",[34,35,37],"h2",{"id":36},"what-cncf-graduated-means-a-seal-of-trust","What CNCF Graduated Means: A Seal of Trust",[11,39,40,41,46],{},"To achieve ",[20,42,45],{"href":43,"rel":44},"https:\u002F\u002Fwww.cncf.io\u002Fprojects\u002F",[24],"Graduated status from CNCF",", projects must meet rigorous criteria:",[48,49,50,57,63,69,75],"ul",{},[51,52,53,56],"li",{},[15,54,55],{},"Broad Production Adoption",": Proven usage in production at multiple organizations",[51,58,59,62],{},[15,60,61],{},"Healthy Community",": Diverse contributors and active development",[51,64,65,68],{},[15,66,67],{},"Security Audit",": Completion of third-party security audits",[51,70,71,74],{},[15,72,73],{},"Governance",": Clear project governance and decision-making processes",[51,76,77,80],{},[15,78,79],{},"Mature Documentation",": Comprehensive documentation and user guides",[11,82,83,84,89],{},"According to the ",[20,85,88],{"href":86,"rel":87},"https:\u002F\u002Fwww.cncf.io\u002Freports\u002Fcncf-annual-report-2025\u002F",[24],"CNCF Annual Report 2025",", CNCF hosts over 230 projects with more than 300,000 contributors. Graduated projects represent the elite tier.",[11,91,92,97],{},[20,93,96],{"href":94,"rel":95},"https:\u002F\u002Fwww.hexabase.com\u002Fproduct\u002Fcaptain-ai\u002F",[24],"Captain.AI"," leverages these CNCF Graduated projects to provide AI-powered assistance for building and operating Kubernetes environments.",[34,99,101],{"id":100},"orchestration-and-runtime","Orchestration and Runtime",[103,104,106],"h3",{"id":105},"kubernetes-the-container-orchestration-standard","Kubernetes -- The Container Orchestration Standard",[11,108,109,114],{},[20,110,113],{"href":111,"rel":112},"https:\u002F\u002Fkubernetes.io\u002F",[24],"Kubernetes"," was CNCF's first Graduated project and is the de facto standard for container orchestration. It automates the deployment, scaling, and management of containerized applications.",[103,116,118],{"id":117},"containerd-cri-o-container-runtimes","containerd \u002F CRI-O -- Container Runtimes",[11,120,121,126,127,132],{},[20,122,125],{"href":123,"rel":124},"https:\u002F\u002Fcontainerd.io\u002F",[24],"containerd"," and ",[20,128,131],{"href":129,"rel":130},"https:\u002F\u002Fcri-o.io\u002F",[24],"CRI-O"," are lightweight container runtimes for Kubernetes. As CRI (Container Runtime Interface)-compliant runtimes independent of the Docker daemon, they are widely adopted in production.",[103,134,136],{"id":135},"crossplane-infrastructure-orchestration","Crossplane -- Infrastructure Orchestration",[11,138,139,144],{},[20,140,143],{"href":141,"rel":142},"https:\u002F\u002Fwww.crossplane.io\u002F",[24],"Crossplane"," achieved Graduated status in 2025, enabling declarative management of cloud infrastructure through the Kubernetes API. AWS, GCP, and Azure resources can be managed uniformly as Kubernetes manifests.",[103,146,148],{"id":147},"knative-serverless","Knative -- Serverless",[11,150,151,156,157,162],{},[20,152,155],{"href":153,"rel":154},"https:\u002F\u002Fknative.dev\u002F",[24],"Knative"," achieved Graduated status in October 2025. According to the ",[20,158,161],{"href":159,"rel":160},"https:\u002F\u002Fwww.cncf.io\u002Fannouncements\u002F2025\u002F10\u002F08\u002Fcloud-native-computing-foundation-announces-knatives-graduation\u002F",[24],"CNCF announcement",", it provides an event-driven application layer on Kubernetes with native support for auto-scaling and scale-to-zero.",[103,164,166],{"id":165},"keda-event-driven-autoscaling","KEDA -- Event-Driven Autoscaling",[11,168,169,174],{},[20,170,173],{"href":171,"rel":172},"https:\u002F\u002Fkeda.sh\u002F",[24],"KEDA"," is an event-driven autoscaler that scales Kubernetes workloads based on external event sources such as Kafka, RabbitMQ, and Azure Queue.",[103,176,178],{"id":177},"kubeedge-edge-computing","KubeEdge -- Edge Computing",[11,180,181,186],{},[20,182,185],{"href":183,"rel":184},"https:\u002F\u002Fkubeedge.io\u002F",[24],"KubeEdge"," extends Kubernetes to edge devices, enabling unified management across cloud and edge environments.",[11,188,189,192],{},[20,190,31],{"href":29,"rel":191},[24]," benefits from these orchestration technologies as a lightweight K3s-based Kubernetes platform spanning edge to cloud.",[34,194,196],{"id":195},"observability","Observability",[103,198,200],{"id":199},"prometheus-the-metrics-collection-standard","Prometheus -- The Metrics Collection Standard",[11,202,203,208,209,213],{},[20,204,207],{"href":205,"rel":206},"https:\u002F\u002Fprometheus.io\u002F",[24],"Prometheus"," is the foundation of Kubernetes monitoring with its pull-based metrics collection and PromQL query language. See our ",[20,210,212],{"href":29,"rel":211},[24],"Prometheus monitoring guide"," on this blog for details.",[103,215,217],{"id":216},"opentelemetry-unified-observability","OpenTelemetry -- Unified Observability",[11,219,220,225],{},[20,221,224],{"href":222,"rel":223},"https:\u002F\u002Fopentelemetry.io\u002F",[24],"OpenTelemetry"," is a vendor-neutral framework that unifies traces, metrics, and logs. Supported by over 90 observability vendors, it has established itself as the industry standard.",[103,227,229],{"id":228},"jaeger-distributed-tracing","Jaeger -- Distributed Tracing",[11,231,232,237],{},[20,233,236],{"href":234,"rel":235},"https:\u002F\u002Fwww.jaegertracing.io\u002F",[24],"Jaeger"," is a distributed tracing backend for microservice environments. It visualizes the complete picture of requests and helps identify performance bottlenecks.",[103,239,241],{"id":240},"fluentd-log-collection","Fluentd -- Log Collection",[11,243,244,249],{},[20,245,248],{"href":246,"rel":247},"https:\u002F\u002Fwww.fluentd.org\u002F",[24],"Fluentd"," is an open-source data collector providing a unified logging layer. With over 500 plugins, it supports diverse data sources and destinations.",[11,251,252,255],{},[20,253,96],{"href":94,"rel":254},[24]," integrates and analyzes data from these observability tools using AI, automating early incident detection and response.",[34,257,259],{"id":258},"networking-and-service-mesh","Networking and Service Mesh",[103,261,263],{"id":262},"cilium-ebpf-based-networking","Cilium -- eBPF-Based Networking",[11,265,266,271],{},[20,267,270],{"href":268,"rel":269},"https:\u002F\u002Fcilium.io\u002F",[24],"Cilium"," is a CNCF Graduated project that leverages eBPF technology to provide L3-L7 networking, security, and observability. It features performance beyond iptables and an identity-based security model.",[103,273,275],{"id":274},"envoy-service-proxy","Envoy -- Service Proxy",[11,277,278,283],{},[20,279,282],{"href":280,"rel":281},"https:\u002F\u002Fwww.envoyproxy.io\u002F",[24],"Envoy"," is a high-performance service proxy developed at Lyft. Used as the foundation for Istio and many service meshes, it provides L7 traffic management, load balancing, and observability.",[103,285,287],{"id":286},"coredns-service-discovery","CoreDNS -- Service Discovery",[11,289,290,295],{},[20,291,294],{"href":292,"rel":293},"https:\u002F\u002Fcoredns.io\u002F",[24],"CoreDNS"," is the default DNS server for Kubernetes. Its plugin-based architecture enables flexible customization of DNS-based service discovery.",[103,297,299],{"id":298},"istio-service-mesh","Istio -- Service Mesh",[11,301,302,307],{},[20,303,306],{"href":304,"rel":305},"https:\u002F\u002Fistio.io\u002F",[24],"Istio"," is a service mesh that manages communication between microservices. It transparently provides traffic management, security (mTLS), and observability.",[103,309,311],{"id":310},"linkerd-lightweight-service-mesh","Linkerd -- Lightweight Service Mesh",[11,313,314,319],{},[20,315,318],{"href":316,"rel":317},"https:\u002F\u002Flinkerd.io\u002F",[24],"Linkerd"," is a lightweight, simple service mesh for Kubernetes. It delivers mTLS, metrics, and retries with minimal configuration.",[34,321,323],{"id":322},"security-and-compliance","Security and Compliance",[103,325,327],{"id":326},"cert-manager-automated-tls-certificate-management","cert-manager -- Automated TLS Certificate Management",[11,329,330,335],{},[20,331,334],{"href":332,"rel":333},"https:\u002F\u002Fcert-manager.io\u002F",[24],"cert-manager"," automates TLS certificate issuance and renewal on Kubernetes. It supports multiple certificate authorities including Let's Encrypt and HashiCorp Vault.",[103,337,339],{"id":338},"falco-runtime-security","Falco -- Runtime Security",[11,341,342,347],{},[20,343,346],{"href":344,"rel":345},"https:\u002F\u002Ffalco.org\u002F",[24],"Falco"," is a cloud-native runtime security engine that monitors Linux kernel system calls and detects abnormal behavior at the container and host level in real time.",[103,349,351],{"id":350},"opa-open-policy-agent-policy-engine","OPA (Open Policy Agent) -- Policy Engine",[11,353,354,359],{},[20,355,358],{"href":356,"rel":357},"https:\u002F\u002Fwww.openpolicyagent.org\u002F",[24],"Open Policy Agent"," is a general-purpose policy engine for cloud-native environments. It handles Kubernetes admission control, API authorization, and data filtering.",[103,361,363],{"id":362},"kyverno-kubernetes-native-policy","Kyverno -- Kubernetes-Native Policy",[11,365,366,371],{},[20,367,370],{"href":368,"rel":369},"https:\u002F\u002Fkyverno.io\u002F",[24],"Kyverno"," is a Kubernetes-native policy engine. Policies are written in YAML without requiring the Rego language, offering an intuitive user experience.",[103,373,375],{"id":374},"spiffe-spire-workload-identity","SPIFFE \u002F SPIRE -- Workload Identity",[11,377,378,383,384,389],{},[20,379,382],{"href":380,"rel":381},"https:\u002F\u002Fspiffe.io\u002F",[24],"SPIFFE"," defines the workload identity standard, while ",[20,385,388],{"href":386,"rel":387},"https:\u002F\u002Fspiffe.io\u002Fspire\u002F",[24],"SPIRE"," is its implementation. Together they automate service-to-service authentication, providing the foundation for zero-trust security.",[103,391,393],{"id":392},"in-toto-tuf-software-supply-chain","in-toto \u002F TUF -- Software Supply Chain",[11,395,396,126,401,406],{},[20,397,400],{"href":398,"rel":399},"https:\u002F\u002Fin-toto.io\u002F",[24],"in-toto",[20,402,405],{"href":403,"rel":404},"https:\u002F\u002Ftheupdateframework.io\u002F",[24],"TUF (The Update Framework)"," are frameworks that guarantee software supply chain integrity, cryptographically verifying each stage from build to deployment.",[34,408,410],{"id":409},"storage-registry-and-ci-cd","Storage, Registry, and ci-cd",[103,412,414],{"id":413},"rook-cubefs-cloud-native-storage","Rook \u002F CubeFS -- Cloud-Native Storage",[11,416,417,422,423,428],{},[20,418,421],{"href":419,"rel":420},"https:\u002F\u002Frook.io\u002F",[24],"Rook"," is a storage orchestrator that automates Ceph management on Kubernetes. ",[20,424,427],{"href":425,"rel":426},"https:\u002F\u002Fcubefs.io\u002F",[24],"CubeFS"," is a large-scale distributed file system.",[103,430,432],{"id":431},"harbor-dragonfly-container-registry","Harbor \u002F Dragonfly -- Container Registry",[11,434,435,440,441,446],{},[20,436,439],{"href":437,"rel":438},"https:\u002F\u002Fgoharbor.io\u002F",[24],"Harbor"," is an enterprise container registry. ",[20,442,445],{"href":443,"rel":444},"https:\u002F\u002Fd7y.io\u002F",[24],"Dragonfly"," is a P2P-based container image distribution system.",[103,448,450],{"id":449},"argo-ci-cd-and-gitops","Argo -- ci-cd and GitOps",[11,452,453,458],{},[20,454,457],{"href":455,"rel":456},"https:\u002F\u002Fargoproj.github.io\u002F",[24],"Argo"," is a Kubernetes-native ci-cd toolset comprising Argo CD (GitOps), Argo Workflows (workflow engine), and Argo Rollouts (progressive delivery).",[103,460,462],{"id":461},"flux-gitops","Flux -- GitOps",[11,464,465,470],{},[20,466,469],{"href":467,"rel":468},"https:\u002F\u002Ffluxcd.io\u002F",[24],"Flux"," is a GitOps tool that automatically synchronizes Git repository state to clusters. It is one of the two major GitOps projects alongside Argo CD.",[103,472,474],{"id":473},"helm-package-manager","Helm -- Package Manager",[11,476,477,482],{},[20,478,481],{"href":479,"rel":480},"https:\u002F\u002Fhelm.sh\u002F",[24],"Helm"," is the package manager for Kubernetes applications. Using templates called Charts, it simplifies the distribution and deployment of complex applications.",[103,484,486],{"id":485},"additional-projects","Additional Projects",[48,488,489,497,505,513,521],{},[51,490,491,496],{},[20,492,495],{"href":493,"rel":494},"https:--dapr.io-",[24],"dapr"," -- distributed application runtime",[51,498,499,504],{},[20,500,503],{"href":501,"rel":502},"https:\u002F\u002Fcloudevents.io\u002F",[24],"CloudEvents"," -- Event data specification standard",[51,506,507,512],{},[20,508,511],{"href":509,"rel":510},"https:\u002F\u002Fetcd.io\u002F",[24],"etcd"," -- Distributed key-value store (Kubernetes' foundation)",[51,514,515,520],{},[20,516,519],{"href":517,"rel":518},"https:\u002F\u002Ftikv.org\u002F",[24],"TiKV"," -- Distributed transactional key-value store",[51,522,523,528],{},[20,524,527],{"href":525,"rel":526},"https:\u002F\u002Fvitess.io\u002F",[24],"Vitess"," -- MySQL scaling solution",[34,530,532],{"id":531},"kubo-and-the-cncf-ecosystem","Kubo and the CNCF Ecosystem",[11,534,535,538],{},[20,536,31],{"href":29,"rel":537},[24]," is a lightweight K3s-based Kubernetes platform with strong affinity for CNCF Graduated projects:",[540,541,542,555],"table",{},[543,544,545],"thead",{},[546,547,548,552],"tr",{},[549,550,551],"th",{},"Category",[549,553,554],{},"Kubo Usage",[556,557,558,569,579,589,599,609,619],"tbody",{},[546,559,560,566],{},[561,562,563],"td",{},[15,564,565],{},"Orchestration",[561,567,568],{},"K3s (Kubernetes-conformant) as the base",[546,570,571,576],{},[561,572,573],{},[15,574,575],{},"Monitoring",[561,577,578],{},"Prometheus + Grafana for standard monitoring",[546,580,581,586],{},[561,582,583],{},[15,584,585],{},"Networking",[561,587,588],{},"Cilium \u002F CoreDNS for high-performance networking",[546,590,591,596],{},[561,592,593],{},[15,594,595],{},"Security",[561,597,598],{},"cert-manager for automated TLS management",[546,600,601,606],{},[561,602,603],{},[15,604,605],{},"ci-cd",[561,607,608],{},"Argo CD \u002F Flux for GitOps deployments",[546,610,611,616],{},[561,612,613],{},[15,614,615],{},"Storage",[561,617,618],{},"Rook\u002FCeph for persistent storage",[546,620,621,626],{},[561,622,623],{},[15,624,625],{},"Registry",[561,627,628],{},"Harbor for private container registry",[11,630,631,632,637],{},"As the ",[20,633,636],{"href":634,"rel":635},"https:\u002F\u002Fwww.cncf.io\u002Fblog\u002F2026\u002F02\u002F09\u002Fwhat-cncf-project-velocity-in-2025-reveals-about-cloud-natives-future\u002F",[24],"CNCF project velocity report"," shows, the cloud-native ecosystem is growing at an accelerating pace. Kubo is a platform that maximizes the benefits of this ecosystem.",[11,639,640,641,644],{},"Combining ",[20,642,96],{"href":94,"rel":643},[24]," with Kubo creates an intelligent platform where AI assists in building and operating production environments that leverage CNCF Graduated projects.",[34,646,648],{"id":647},"conclusion","Conclusion",[11,650,651],{},"As of 2025, there are 36 CNCF Graduated projects covering every area of cloud-native technology:",[653,654,655,661,667,673,679],"ol",{},[51,656,657,660],{},[15,658,659],{},"Graduated status"," certifies broad production adoption, security audits, and mature governance",[51,662,663,666],{},[15,664,665],{},"Centered on Kubernetes",", the ecosystem covers orchestration, observability, networking, and security",[51,668,669,672],{},[15,670,671],{},"New 2025 graduates"," Crossplane and Knative strengthen IaC and serverless capabilities",[51,674,675,678],{},[15,676,677],{},"Cilium and OpenTelemetry"," demonstrate the maturation of innovative eBPF and observability projects",[51,680,681,684],{},[15,682,683],{},"The entire ecosystem"," works together organically to form a comprehensive cloud-native foundation",[11,686,687,690,691,694],{},[20,688,31],{"href":29,"rel":689},[24]," seamlessly integrates with these CNCF Graduated projects on a K3s base, delivering enterprise-grade cloud-native environments. If you are interested in leveraging the CNCF ecosystem, explore ",[20,692,31],{"href":29,"rel":693},[24],".",[11,696,697,698,701,702,694],{},"For AI-powered cloud-native operations automation, see ",[20,699,96],{"href":94,"rel":700},[24],". For consultations, reach out through our ",[20,703,706],{"href":704,"rel":705},"https:\u002F\u002Fwww.hexabase.com\u002Fcontact-us\u002F",[24],"contact page",{"title":708,"searchDepth":709,"depth":709,"links":710},"",2,[711,712,721,727,734,742,750,751],{"id":36,"depth":709,"text":37},{"id":100,"depth":709,"text":101,"children":713},[714,716,717,718,719,720],{"id":105,"depth":715,"text":106},3,{"id":117,"depth":715,"text":118},{"id":135,"depth":715,"text":136},{"id":147,"depth":715,"text":148},{"id":165,"depth":715,"text":166},{"id":177,"depth":715,"text":178},{"id":195,"depth":709,"text":196,"children":722},[723,724,725,726],{"id":199,"depth":715,"text":200},{"id":216,"depth":715,"text":217},{"id":228,"depth":715,"text":229},{"id":240,"depth":715,"text":241},{"id":258,"depth":709,"text":259,"children":728},[729,730,731,732,733],{"id":262,"depth":715,"text":263},{"id":274,"depth":715,"text":275},{"id":286,"depth":715,"text":287},{"id":298,"depth":715,"text":299},{"id":310,"depth":715,"text":311},{"id":322,"depth":709,"text":323,"children":735},[736,737,738,739,740,741],{"id":326,"depth":715,"text":327},{"id":338,"depth":715,"text":339},{"id":350,"depth":715,"text":351},{"id":362,"depth":715,"text":363},{"id":374,"depth":715,"text":375},{"id":392,"depth":715,"text":393},{"id":409,"depth":709,"text":410,"children":743},[744,745,746,747,748,749],{"id":413,"depth":715,"text":414},{"id":431,"depth":715,"text":432},{"id":449,"depth":715,"text":450},{"id":461,"depth":715,"text":462},{"id":473,"depth":715,"text":474},{"id":485,"depth":715,"text":486},{"id":531,"depth":709,"text":532},{"id":647,"depth":709,"text":648},"2026-05-27","A complete guide to all 36 CNCF Graduated projects. From Kubernetes and Prometheus to Cilium and Crossplane, explore production-proven OSS by category.","md","en",{},true,"\u002Fblog\u002Fen\u002Fcncf-graduated-projects-2025-overview",{"title":5,"description":753},"blog\u002Fen\u002Fcncf-graduated-projects-2025-overview",[762,763,764,765,766,767,768,769],"cncf","graduated","cloud-native","oss","kubernetes","prometheus","cilium","opentelemetry","lKkv4jT3uLgn9vuxsahr3MygCpnGpu5QoEnBSUSMkFk",[772,781,791,800,809,816],{"path":773,"title":774,"description":775,"date":776,"tags":777},"\u002Fblog\u002Fen\u002Fcncf-graduated-project-oss-selection-criteria","Should You Trust the CNCF 'Graduated' Badge? What TOC Public Meetings Reveal About the Unglamorous Reality of the Review Process","The CNCF 'Graduated' badge is not a safety certificate. Drawing on TOC public meeting notes and official documentation, this article explains the real review process behind Sandbox, Incubating, and Graduated status, and how to judge OSS maturity before adopting it in production.","2026-07-23",[762,766,778,765,779,780],"k3s","managed-kubernetes","governance",{"path":782,"title":783,"description":784,"date":752,"tags":785},"\u002Fblog\u002Fen\u002Fcilium-ebpf-kubernetes-networking","Revolutionizing Kubernetes Networking with Cilium and eBPF","Discover how Cilium uses eBPF to transform Kubernetes networking. Compare with traditional CNIs, explore L7 policies, and leverage Hubble for observability.",[768,786,766,787,788,762,789,790],"ebpf","networking","cni","security","hubble",{"path":792,"title":793,"description":794,"date":752,"tags":795},"\u002Fblog\u002Fen\u002Fgrafana-dashboard-kubernetes-observability","Grafana Dashboard Design: Kubernetes Observability in Practice","Design effective Grafana dashboards for Kubernetes observability using the RED method, Four Golden Signals, and production-proven best practices.",[796,766,195,797,762,767,798,799],"grafana","dashboard","monitoring","visualization",{"path":801,"title":802,"description":803,"date":752,"tags":804},"\u002Fblog\u002Fen\u002Fopentelemetry-distributed-tracing","Implementing Distributed Tracing with OpenTelemetry: From Basics to Production","A practical guide to distributed tracing with OpenTelemetry. Covers Collector setup, auto-instrumentation, Jaeger\u002FTempo integration, and Kubernetes best practices.",[769,805,766,762,195,806,807,808],"distributed-tracing","jaeger","tempo","microservices",{"path":810,"title":811,"description":812,"date":752,"tags":813},"\u002Fblog\u002Fen\u002Fprometheus-monitoring-kubernetes-guide","The Complete Guide to Monitoring Kubernetes with Prometheus","Master Kubernetes monitoring with Prometheus: from installation and PromQL queries to Alertmanager configuration and production best practices.",[767,766,798,762,195,814,815],"promql","alertmanager",{"path":817,"title":818,"description":819,"date":820,"tags":821},"\u002Fblog\u002Fen\u002Fk3s-ai-agent-opentelemetry-observability","Your AI Agent Is Running on K3s, But You Have No Idea What It's Doing. Tracing Non-Deterministic Execution with OpenTelemetry","MCP servers and AI agents on K3s don't reveal why they called a given tool just from logs. Here's how OpenTelemetry's GenAI semantic conventions make non-deterministic execution observable.","2026-09-06",[778,766,769,805,822,823],"genai","telemetry",1792488897044]