Infrastructure, SRE & Incident Operations

Adds a Kubernetes-native agent control plane for cluster automation experiments that still need operator boundaries.

Lets operators inspect and act on Kubernetes resources through MCP instead of brittle one-off kubectl prompts.

Covers cluster-level MCP operations for routine checks, triage, and controlled remediation.

Extends the cluster-ops workflow to Kubernetes and OpenShift environments where enterprise platform teams need consistent runbooks.

Pulls Kubernetes and cloud signals into an incident investigation loop instead of leaving responders to correlate dashboards manually.

Gives responders a fast multi-pod log tailing step during label-scoped outages.

Turns CrashLoopBackOff and pod failure evidence into a focused diagnosis packet for responders.

Provides a fast terminal cockpit for cluster inspection during normal operations and incidents.

Checks Kubernetes workload readiness and policy hygiene before manifest mistakes reach production.

Prevents configuration automation failures by linting playbooks before they run against infrastructure.

Automates GitOps sync checks and actions for teams running ArgoCD-backed production deploys.

Makes Helm upgrades reviewable by surfacing chart diffs before cluster changes land.

Coordinates many Helm releases from a single desired state so platform drift is visible before apply.

Runs declarative Kubernetes tests before operator or manifest changes become incident fuel.

Turns Terraform plan output into an infrastructure-review artifact before risky changes merge.

Detects drift between declared Terraform state and real infrastructure before incidents or surprise costs appear.

Covers backup, restore, retention, and failure notification as an auditable runbook rather than an ad hoc database chore.

Centralizes scheduled backups and retention across databases for smaller platform teams.

Adds encrypted file and repository backup workflows for production recovery planning.

Gives infrastructure teams immediate host and service telemetry for alerting and triage.

Provides logs, metrics, and traces for production reliability investigations without a proprietary-only observability stack.

Adds a cloud-native observability backend for logs, metrics, traces, and incident search.

Routes legacy SMTP-only alerts into modern notification channels so incidents reach the right responders.

Connects Datadog signals into agent workflows for incident context, monitor review, and dashboard triage.

Adds run-level tracing for agent systems that SRE teams need to debug after failures or noisy automation.

Connects LLM and agent traces to OpenTelemetry-style incident investigation workflows.

Adds OpenTelemetry-style traces, metrics, and evaluations for production LLM and agent applications that SRE teams need to debug and monitor.

Profiles CPU, GPU, and memory hotspots in Python agent services, tool servers, or retrieval pipelines before reliability tuning.

Adds monitoring and evaluation coverage for production LLM workflows that need regression signals during incidents.

Records system-level behavior from agent commands during incidents or postmortems, giving SRE teams evidence beyond agent transcripts.