Tools every DevOps engineer should know

September 25, 2026 · 15 min read

DevOps engineers live at the intersection of APIs, configs, and clusters. The tools that matter parse JSON reliably, reproduce HTTP calls, and decode tokens without leaking them into ticket systems. This list mixes industry staples with browser utilities that fill gaps when SSH is not available.

CLI essentials

• curl - reproduce webhooks and health checks from any jump host.
• jq - slice JSON logs and Helm values in pipelines.
• kubectl - inspect resources; pair with -o json and jq.
• terraform - plan/apply with versioned state; validate JSON policies.

# Extract failing pod status
kubectl get pod -n prod -o json | jq '.items[] | select(.status.phase!="Running") | .metadata.name'

Observability stack

Metrics (Prometheus), logs (Loki/ELK), and traces (Jaeger/Tempo) form the triad. Learn label cardinality rules before you explode metric series. Correlate trace IDs with request IDs in API gateways.

JSON and config data

Invalid JSON breaks Helm charts, Lambda env vars, and GitOps repos. Keep a validator bookmarked for quick checks when reviewing MRs. Compare tools help when two clusters should have the same ConfigMap shape but drifted.

Secrets and tokens

Decode JWT service account tokens only on secure machines. Rotate keys through your secret manager, not Slack. Browser decoders are for staging tokens during pairing, not for production credential inspection on shared screens.

Incident toolkit

During outages: status page, runbooks, feature flags, and a known-good curl for the critical path. Paste formatted JSON from gateways into the bridge doc so everyone sees the same error body. Roll back first; root-cause with compare tools second.

Related: Free API debugging utilities

Browse all tools