orca is a Kubernetes terminal cockpit inspired by k9s, built in Rust with ratatui.

• Async TUI runtime with tokio + typed Kubernetes client (kube, k8s-openapi)
• Powerline-style header/footer with compact context, scope, and status data
• Multi-view workflow (view slots) for fast context switching without losing state
• Vim-style navigation and command/jump/filter modes
• Single main pane that can show:
  • resource table
  • dashboard overview (o)
  • syntax-highlighted details (d)
  • logs/output overlay
  • embedded pod shell overlay (s, :shell, :ssh, :bash)
  • pod container picker
• Watch-based refresh for mapped resources, with periodic refresh fallback
• Context, cluster, and user switching from kubeconfig
• Context/user/cluster catalog overlays (:ctx, :usr, :cluster without args)
• Namespace scoping (--namespace, --all-namespaces, :ns, :all-ns)
• Pod/service port-forward management with live PF indicators
• YAML/JSON syntax highlighting in details view
• DevOps tool overlays for Argo CD, Helm, Terraform, Ansible, Docker, OpenShift, and Kustomize
• Git repo toolkit for fetching/caching repos and applying/exporting content (:git, :repo)
• Fleet pulse snapshot (:pulses) and resource relationship trace (:xray)
• Persistent incident snapshot counters in footer glance (warnings, crashloops, not-ready nodes)
• Read-only safety mode (:readonly on|off|toggle, ORCA_READONLY=1)
• Runtime aliases/plugins/hotkeys from YAML config with automatic reload

• Pods
• CronJobs
• DaemonSets
• Deployments
• ReplicaSets
• ReplicationControllers
• StatefulSets
• Jobs
• Services
• Ingresses
• IngressClasses
• ConfigMaps
• PersistentVolumeClaims
• Secrets
• StorageClasses
• PersistentVolumes
• ServiceAccounts
• Roles
• RoleBindings
• ClusterRoles
• ClusterRoleBindings
• NetworkPolicies
• Nodes
• Events
• Namespaces
• CRD (custom resources + CRD catalog)

• Access to a Kubernetes cluster ($KUBECONFIG or in-cluster config)
• kubectl in PATH for subprocess actions:
  • :exec
  • :shell / :ssh / :bash
  • :edit
  • :port-forward
• Optional but recommended: metrics-server for richer CPU/RAM dashboard data
• Optional DevOps toolchain for overlays:
  • argocd, helm, terraform, ansible-playbook, docker, oc, kustomize, git

orca uses $KUBE_EDITOR for :edit; if unset, it forwards $EDITOR to kubectl. Set ORCA_READONLY=1 to start in safety mode where mutating actions are blocked. Set ORCA_CONFIG=/path/to/orca.yaml to pin a specific runtime config file.

cargo run --release -- --all-namespaces

cargo run --release -- --namespace default --refresh-ms 1500

• --refresh-ms <ms>: refresh interval in milliseconds (minimum enforced at runtime: 500)
• -n, --namespace <name>: start in a specific namespace
• -A, --all-namespaces: start with all namespaces
• --log-filter <level>: tracing filter (default: info)

• Enter is drill-down, not details:
  • Namespaces -> Pods (sets namespace scope)
  • Pods -> Containers (container picker)
  • Deployments/DaemonSets/StatefulSets/ReplicaSets/ReplicationControllers/Jobs/CronJobs -> Pods
  • Services -> Pods
• d opens details mode for the selected row
• Esc goes back one step (shell/logs -> containers -> previous flow/root)
• o toggles overview dashboard in the main pane

• Left / Right: previous/next resource tab
• j / k, Up / Down: move selection
• gg / G: top / bottom
• Ctrl+u / Ctrl+d, PageUp / PageDown: page scroll
• /: filter mode
• :: command mode
• >: jump mode
• Tab (input modes): autocomplete
• Up / Down or Ctrl+p / Ctrl+n (input modes): autocomplete selection
• Enter (or terminal fallbacks Ctrl+m / Ctrl+j in input mode): submit input
• l: logs for selected pod/container
• Shift+L: previous/related logs (workload/service aware)
• s: open embedded shell (/bin/sh) in selected pod (inside ORCA)
• e: edit selected resource
• p: prefill :port-forward command
• d: open details view
• o: open/close overview
• Tab (normal mode): toggle focus (table/details when details mode is active)
• y / n: confirm or cancel pending actions
• ?: help modal
• q: quit

• Ctrl+1..9 switches/creates view slots
• 1..9 also switches view slots in normal mode
• Ctrl+Shift+1..9 switches view slots in input modes
• Ctrl+Alt+0..9 deletes a view slot
• View state is preserved per slot (tab, scope, filter, overlays, selection)

Supported commands:

• :q, :quit, :exit
• :refresh (:reload, :r)
• :ctx <context> (:context, :use-context)
• :cluster <cluster-name-or-server> (:cl)
• :user <user> (:usr)
• :contexts, :clusters, :users
• :all-ns (:all, :allns, :all-namespaces)
• :ns / :namespace / :namespaces
• :ns <namespace>
• :<resource> (switch tab by alias)
• :<resource> <filter>
• :<resource> <namespace>/<name>
• :filter <query>
• :clear
• :logs
• :edit (:e)
• :delete (:del) (confirmation required)
• :restart (Deployments/StatefulSets, confirmation required)
• :scale <replicas> (Deployments/StatefulSets, immediate)
• :exec <command...> (Pods tab)
• :shell [container] [shell]
• :ssh [container] [shell]
• :bash
• :pf <local>:<remote> (:port-forward)
• :crd <name|kind|plural> (:custom)
• :crd-refresh
• :help
• :ops, :tools
• :readonly on|off|toggle (:ro)
• :config (shows loaded config source, aliases, plugins, hotkeys)
• :alerts (:alert) high-signal incident snapshot
• :pulses (:pulse)
• :xray (:xr, :x) on selected row (or explicit target)
• :argocd [app-name]
• :helm [release]
• :tf (:terraform)
• :ansible (:ans)
• :docker
• :rbac [subject] (uses kubectl auth can-i --list, optional --as)
• :who-can <verb> <resource> [namespace] (uses kubectl-who-can plugin or kubectl who-can)
• :oc (:openshift)
• :kustomize [path]
• :git / :repo (toolkit catalog)
• :git fetch <url-or-repo> [ref]
• :git files <url-or-repo> [path]
• :git show <url-or-repo> <path>
• :git export <url-or-repo> <source-path> [destination]
• :git apply <url-or-repo> <path>
• :plugin <name> [args...] (:plug) runs configured plugin command

Compatibility command:

• :tab <resource> [filter-or-target]

• Supports the same context/cluster/user and resource aliases for fast navigation
• Supports DevOps overlays (>tools, >argocd, >helm, >tf, >ansible, >docker, >rbac, >who-can, >oc, >kustomize, >git)
• Supports observability overlays (>pulses, >xray)
• Supports incident overlays (>alerts)
• Supports config/plugin actions (>config, >plugin <name> ...)
• Supports namespaced targets (for example >po my-ns/my-pod)
• Supports fuzzy jump by resource name/namespace when no explicit alias is provided
• Resets to the current flow root before executing jump selection

Short aliases accepted in : and > include:

• po, cj, ds, deploy, rs, rc, sts, job
• svc, ing, ingclass, cm, pvc, secret, sc, pv
• sa, role, rb, crole, crb, np, node, event, ns, crd

Long names (pods, deployments, services, etc.) are also supported.

• :delete and :restart are guarded by confirmation (y/n, Enter also confirms)
• :scale executes immediately and refreshes the active resource table
• l/:logs are pod/container log focused
• Shift+L resolves related pod logs for workload/service resources
• s / :shell / :ssh / :bash open an embedded shell overlay (sh pane label); Esc closes it
• Port-forward sessions are tracked and shown in:
  • PF table column for Pods/Services
  • header badge for selected resource

ORCA can load aliases and plugins from:

• ORCA_CONFIG (if set)
• ./orca.yaml, ./orca.yml, ./.orca.yaml
• $HOME/.config/orca/config.yaml

Example:

aliases:
  d: "deployments"
  sys: "ns kube-system"

plugins:
  - name: "describe-pod"
    command: "kubectl"
    args: ["describe", "pod", "{name}", "-n", "{namespace}"]
    description: "Describe currently selected pod"
    mutating: false
    timeout_secs: 20
    retries: 1

hotkeys:
  - key: "ctrl+shift+p"
    command: "pulses"
    description: "Open fleet pulses"
  - key: "ctrl+shift+x"
    command: "xray"
    description: "Open xray for selected row"
  - key: "ctrl+shift+g"
    command: "po kube-system/coredns"
    jump: true
    description: "Jump directly to coredns pod"

Supported placeholders in plugin args:

• {name}, {namespace}, {target}, {resource}
• {context}, {cluster}, {user}, {scope}
• {all_namespaces}, {args}
• {extra} to splice all user-supplied plugin args

Plugin runtime profile fields:

• timeout_secs: command timeout per attempt (clamped 1..300, default 20)
• retries: retry count after the first failed attempt (clamped 0..5, default 0)

• src/main.rs: runtime loop, event handling, refresh/watch orchestration
• src/app.rs: state machine, mode handling, command parser, drill-down flow
• src/input.rs: key mapping by mode
• src/k8s.rs: Kubernetes API gateway, table builders, actions, metrics, discovery
• src/config.rs: runtime YAML config loader/watcher (aliases + plugins)
• src/ui.rs: ratatui rendering, powerline bars, dashboard, syntax highlighting
• src/model.rs: shared tab/data models
• src/cli.rs: CLI argument definitions