New to this? Start with Foundations. Already running systems? Jump to Agent operations. Everything points to upstream material you can read and use.
The index
Search by problem, system or practice. Use Foundations for first-line IT, or Agent operations for the sharp edge. Every card goes upstream.
A hands-on class for the shell, command-line tools, Git and more. Start here if the terminal still feels like a separate profession.
missing.csail.mit.edu
A free book on files, permissions, processes and shell scripts. Read a chapter, then practise on a disposable machine.
linuxcommand.org
The free Pro Git book, from commits and branches to remotes, history and recovery. Learn how to make small changes safe to undo.
git-scm.com
Requests, responses, URLs, status codes and headers. The vocabulary behind browsers, APIs and reverse proxies.
developer.mozilla.org
The official hands-on path for containers, images, networking and volumes. Build one small service before touching orchestration.
docs.docker.com
An official tutorial for provisioning Azure with Terraform. Learn the plan, apply, state and destroy loop before modules.
developer.hashicorp.com
A Microsoft Learn path covering Azure architecture and core services. Put names to compute, storage, networking and identity.
learn.microsoft.com
An interactive tutorial for deploying, scaling, updating and debugging a containerised app on Kubernetes.
kubernetes.io
Repository automation written in Markdown. Read the permissions, safety and review model before an agent can open or edit changes.
docs.github.com
A shared naming and event model for model calls and agent workflows. Agree the telemetry shape before building dashboards or alerts.
github.com
A framework for increasing supply chain guarantees. Start with provenance and verification when an agent or pipeline can change the artifact.
slsa.dev
Sign software and verify who or what produced it, without turning long-lived key management into a second project.
docs.sigstore.dev
Repeatable checks for common supply-chain controls. Use the result to decide what an agent may safely touch.
scorecard.dev
Policy-as-code that separates a decision from the service enforcing it. Put deployment, identity and network rules outside the agent.
openpolicyagent.org
The interface between a model and its tools. Read the transport, lifecycle and security sections before wiring in real permissions.
modelcontextprotocol.io
The canonical reference for concepts, tasks and the API. Start with Workloads and Services if you are new.
kubernetes.io
Package manager for Kubernetes. Chart structure, templating functions, and the hook lifecycle you need before writing your own chart.
helm.sh
Terminal UI for cluster triage. Hotkeys for logs, describe, port-forward and shell make it the fastest way to debug a failing pod.
k9scli.io
Just-in-time node provisioning. Consolidates under-used nodes automatically and cuts the node-group sprawl that managed autoscalers leave behind.
karpenter.sh
eBPF-powered networking, observability and security for Kubernetes. Hubble gives you flow-level visibility without a sidecar.
docs.cilium.io
Policy engine for Kubernetes written as plain YAML. Validate, mutate and generate resources without learning a new policy language.
kyverno.io
Bootstrap a cluster from binaries with no installers. It is painful on purpose and shows what the control plane actually does.
github.com
Declarative GitOps continuous delivery for Kubernetes. Read the reconciliation and sync-wave docs before you trust it with production.
argo-cd.readthedocs.io
GitOps with modular controllers, multi-tenancy and first-class OCI artifact support.
fluxcd.io
The four properties a system must have to legitimately be called GitOps: declarative, versioned, pulled automatically, continuously reconciled.
opengitops.dev
Workflow syntax, reusable workflows and OIDC federation to cloud roles cover most teams' automation needs without extra tooling.
docs.github.com
Dependency update automation with grouping, scheduling and lock-file support. The config reference is worth an hour of your time.
docs.renovatebot.com
The open-source fork of Terraform under the Linux Foundation. Drop-in for most workflows, with state encryption and provider forking done properly.
opentofu.org
Provider registry, state semantics and the language reference, including import and moved blocks that make refactors survivable.
developer.hashicorp.com
Module layout, state separation, naming and the question of how big a root module should be, with opinionated examples.
terraform-best-practices.com
Pull-request automation for Terraform. Plan output posted on the PR, apply gated behind an approval comment. Small, boring, effective.
runatlantis.io
Static analysis for Terraform, CloudFormation, Kubernetes manifests and Dockerfiles. Catches the public-bucket class of mistake before review.
checkov.io
One vendor-neutral pipeline for traces, metrics and logs. Start with the Collector and semantic conventions, not the SDK rabbit hole.
opentelemetry.io
The pull-based metrics system that defined modern monitoring. PromQL and the recording-rule docs do most of the heavy lifting.
prometheus.io
Label-indexed log aggregation that stays cheap because it does not full-text index. LogQL is worth learning properly.
grafana.com
Distributed tracing backend with dependency graphs and critical-path analysis. The fastest way to find which service actually ate the latency.
jaegertracing.io
Brendan Gregg's performance observability maps, organised by subsystem. Print the 60-second checklist for the next slow host.
brendangregg.com
One binary scanning containers, filesystems, Git repos, IaC and SBOMs. The least-friction first scanner to drop into a pipeline.
trivy.dev
Secrets management with dynamic credentials, leasing and PKI. Read the auth-method and policy docs before designing your paths.
developer.hashicorp.com
Encrypt values inside YAML, JSON and env files so secrets can live in Git next to the code that uses them, with KMS-backed keys.
getsops.io
Site Reliability Engineering, Workbook and the Runbook, free online. Error budgets, SLIs and blameless postmortems, straight from the source.
sre.google
The annual State of DevOps data behind the four key metrics: lead time, deployment frequency, change failure rate and time to restore.
dora.dev
Chaos engineering on Kubernetes as CRDs. Run pod kills, network partitions, clock skew and I/O faults within a namespace.
chaos-mesh.org
A visual path from Linux and scripting through containers, CI/CD and cloud. Use it as a checklist of gaps, not a syllabus to follow in order.
roadmap.sh
Every cloud-native project grouped by layer. Use it with the CNCF trail map to choose a starting point.
landscape.cncf.io
Still the clearest statement of how a service should behave: config in env, stateless processes, logs as streams, disposable at any moment.
12factor.net
Interactive visual sandbox for rebase, cherry-pick and other Git workflows. Useful if you only ever use merge.
learngitbranching.js.org
Public postmortems from teams who broke production with Kubernetes. The most efficient way to inherit someone else's scar tissue.
k8s.af
No resources match that search. Try a shorter keyword, or clear the category filter.