Scenarios¶
Nineteen walkthroughs, from a first lab on your laptop to a FIPS environment, a private GKE cluster, a DR drill and an AI agent with an MCP server. Each page has numbered steps with exact commands, the output to expect, a "verify it worked" check, a clean-up, and what just happened under the hood. Every page includes agent, MCP and UI routes. The checked manifest maps every command and catalog item to a walkthrough; it does not claim every item was deployed live. Together they document cloudseed’s command surface (coverage matrix below).
Every page states its verification level
- New operational scenarios use local checks and deterministic fixtures. Live cloud deployment is pending without sandbox credentials.
- Verified live on VMware Fusion 13.6: the page's script builds real VMs and runs every command (
CLOUDSEED_LIVE=1 tests/scenarios/run.sh 05). - Verified with --dry-run: cloud pages are rendered and checked with
terraform validate, no account needed; the steps that need credentials are listed by the script and are yours to run. - Verified live (local): the MCP server and the web console run for real in a throw-away home.
tests/test_scenarios_docs.pyparses every command on these pages with cloudseed's own parser, so the docs cannot drift from the CLI.
Pick a scenario¶
-
01 · Your first lab on your laptop
A hardened bastion and a private network on VMware Fusion or Workstation: setup, status, SSH, inventory, destroy.
Beginner · 15 min · $0 · Live
-
02 · Secure AWS landing zone
Multi-AZ VPC, NAT, flow logs, bastion open only to your IP, CloudTrail + GuardDuty, S3 remote state, update-ip, targeted destroy.
Intermediate · 25 min · ≈ $58/mo · Dry-run
-
03 · GCP with a private GKE cluster
Custom VPC, Cloud NAT, OS Login bastion, private GKE with Workload Identity, SSH tunnel for kubectl, node scaling.
Intermediate · 30 min · ≈ $197/mo · Dry-run
-
04 · Azure with private AKS
VNet without default outbound, NSGs, NAT gateway, Defender, private AKS, node pool resizing, prowler CIS scan.
Intermediate · 30 min · ≈ $150/mo · Dry-run
-
05 · Kubernetes on your laptop
RKE2 (or kubeadm) on private VMs, kubeconfig,
cs kubectl/helm/k9s, the current environment, a first app.Intermediate · 35 min · $0 · Live
-
06 · A production platform in one command
ArgoCD, Prometheus + Grafana, Loki, OpenTelemetry, Gateway API + Envoy, cert-manager, secrets; UIs over TLS; a CI template.
Intermediate · 30 min · $0 · Live
-
07 · Data and AI stack
MinIO, a CloudNativePG Postgres, a dev Kafka, Ollama + Open WebUI, remembered
--setvalues, Databricks and Snowflake profiles.Advanced · 40 min · $0 · Live
-
08 · Backups you can trust
Velero with its bucket and identity created for you: back up, break, restore, schedule, and an automated DR drill.
Intermediate · 20 min · $0 · Live
-
09 · Chaos engineering with verdicts
Chaos Mesh experiments with steady-state hypotheses, against a canary and your own Deployment; PASS / FAIL reports.
Intermediate · 20 min · $0 · Live
-
10 · Compliance scans
CIS with kube-bench, NSA / MITRE with kubescape, image CVEs with trivy, OpenSCAP CIS and STIG on hosts, reports.
Intermediate · 25 min · $0 · Live
-
11 · FIPS 140 mode
What one switch changes, what it refuses, the platform's FIPS tiers, and end-to-end verification with
cs scan fips.Advanced · 15-45 min · $0 offline · Dry-run
-
12 · Private access with a VPN
A hardened OpenVPN host with its own PKI (users, profiles, connect, revoke) or a Tailscale subnet router.
Intermediate · 20 min · ≈ $70/mo · Dry-run
-
13 · Day-2 operations
Add and remove nodes, change settings and undo them, inventory and audit trail, troubleshooting, the credential vault.
Intermediate · 30 min · $0 · Live
-
14 · AI agents and MCP
Agentic mode with Claude Code, Codex, Gemini or the built-in agent; skills; redaction; an MCP server for your AI clients.
Beginner · 15 min · $0 · Live (local)
-
15 · Web console and FinOps
The local console (wizard, actions, Activity, explain), then estimates, the cloud bill, OpenCost and a report for an agent.
Beginner · 15 min · $0 · Live (local)
Suggested paths¶
| If you ... | Follow |
|---|---|
| have no cloud account (everything local, $0) | 01 → 05 → 06 → 08 → 09 → 10 → 13 |
| run infrastructure on a cloud | 02 or 03 or 04 → 12 → 06 → 08 → 15 |
| work in a regulated environment | 11 → 10 → 02 → 13 |
| want an AI agent to do it | 14 → 15 → 01 |
Operational readiness walkthroughs¶
| Scenario | Outcome | Verification |
|---|---|---|
| 16 · Health and private networking | current observations and an approved in-cluster egress probe | Local/fixtures; cloud pending |
| 17 · Profiles, specs and guardrails | real topology profiles, portable intent, cost/plan gates and explicit expiry | Local/fixtures; cloud pending |
| 18 · Drift, upgrades and recovery | identity-bound upgrade plans and isolated application recovery checks | Local/fixtures; cloud pending |
| 19 · Acceptance and trusted releases | no-account previews, sandbox lifecycle, keychain and artifact evidence | Local/fixtures; cloud pending |
Choose CLI, agent, MCP or UI and inspect feature coverage.
Every feature, and where to try it¶
| Feature | What you do with it | Scenarios |
|---|---|---|
| Landing zones on AWS, GCP, Azure | VPC / VNet, public, private and data subnets, NAT, flow logs, hardened bastion, security baseline, remote state | 02 · 03 · 04 |
| Local VMware environments | The same shape on Fusion Pro / Workstation Pro through cloudseed's own Terraform provider | 01 · 05 |
| Managed and local Kubernetes | Private EKS / GKE / AKS, RKE2 or kubeadm on VMs, kubeconfig, tunnels, cs kubectl / helm / k9s | 03 · 04 · 05 |
| Node pools | cs node add / remove / scale through the cloud API or Terraform + Ansible | 03 · 04 · 13 |
| Platform catalog | Ten groups of pinned Helm / kustomize items: plan, install, uninstall, --set, --upgrade, UIs over TLS | 06 · 07 · 08 · 09 |
| Platform template gitlab-ci | A CI pipeline (build, test, trivy scan, release, ArgoCD deploy) for your app repository | 06 |
| Managed data services | Databricks and Snowflake connection profiles per environment, vendor CLIs passed through | 07 |
| Disaster recovery (Velero) | Bucket + identity created for you, backup, restore, schedules, the automated cs dr test drill | 08 |
| Chaos engineering | Chaos Mesh experiments with steady-state verdicts and saved reports | 09 |
| Compliance and vulnerability scans | kube-bench CIS, kubescape NSA / MITRE, trivy, OpenSCAP CIS / STIG, prowler cloud CIS, reports | 10 · 04 |
| FIPS 140 mode | FIPS endpoints, images, kernels, keys and ciphers; refusals; cs scan fips | 11 · 10 |
| VPN | OpenVPN with its own PKI, or a Tailscale subnet router | 12 |
| Undo | cs undo: fifteen undo points per environment (at most five of one kind), --global, --id, --drop, from CLI, console, agents and MCP | 13 · 06 · 08 · 14 |
| Audit trail, inventory, troubleshooting | logs/audit.jsonl, full redacted logs, inventory.json, deterministic cs troubleshoot | 13 · 01 · 15 |
| Credential vault | cs creds: hidden input, 0600 file, injected everywhere, stripped from agents | 13 · 02 · 14 |
| Dependencies, container runtime, single binary | cs deps install, cs deps image + --runtime container, cs deps bundle, cs install ... | 02 · 05 · 10 |
Help system and cs explain | Per-command help, topics, generated variable and output references, how anything is implemented | 01 · 02 · 15 |
| FinOps | Estimates before you apply, the provider bill, OpenCost allocation, one report for an agent | 15 · 02 · 03 · 04 |
| Web console | cs enable ui: the wizard, environment actions, platform, resilience, reports, explain panel, Activity | 15 |
| AI agents | Agentic mode (built-in, Claude Code, Codex, Gemini, Grok), context brief, session launchers, skills, redaction, approvals and paged saved-evidence review | 14 |
| MCP server | Every feature as a tool for Claude Code, Claude Desktop, Cursor, VS Code, Codex, Gemini CLI, Windsurf | 14 |
| Access and hardening | cs ssh, cs update-ip, cs provision (Ansible hardening, re-runnable) | 01 · 02 · 12 · 13 |
Run the scenarios yourself¶
The scripts live next to the code in tests/scenarios/. They never touch your ~/.cloudseed unless you ask for a live run:
tests/scenarios/run.sh # all 19: cloud/VMware dry runs plus local operational checks
tests/scenarios/run.sh 02 14 # just these
tests/scenarios/run.sh -v 03 # stream every command's output
CLOUDSEED_LIVE=1 tests/scenarios/run.sh 01 05 06 08
python3 -m unittest tests.test_scenarios_docs
run.sh prints a table (scenario, mode, result, checks, skipped live steps, time). A live run builds real VMs with VMware Fusion Pro / Workstation Pro; the cluster scenarios share the vmware-lab cluster from 05 and destroy it at the end.
Command coverage matrix¶
Every command and sub-command of the CLI, and the scenarios whose steps run it. The tables are generated from the pages and checked by tests/test_scenarios_docs.py, so they are always current and nothing is left out.
New shared operations¶
| Features | Walkthrough |
|---|---|
| Live health, DNS/API/registry diagnostics and active egress probe | 16 |
| Deployment profiles, portable specs, cost/destructive guards and expiry | 17 |
| Drift, safe upgrades and isolated application recovery | 18 |
| Sandbox acceptance, native keychain and release verification | 19 |
Environments¶
| Command | Scenarios that run it |
|---|---|
cs setup aws | 02 · 11 · 12 · 15 |
cs setup gcp | 03 · 11 · 12 |
cs setup azure | 04 |
cs setup vmware | 01 · 05 · 11 · 13 |
cs plan | 02 · 13 · 17 |
cs apply | 02 · 13 · 17 |
cs status | 01 · 02 · 03 · 04 · 13 |
cs output | 01 · 02 · 12 |
cs list | 01 · 13 |
cs inventory | 01 · 02 · 13 |
cs troubleshoot | 01 · 02 · 03 · 05 · 12 · 13 |
cs doctor | 01 · 02 · 03 · 04 · 05 · 13 |
cs update-ip | 02 · 13 |
cs provision | 12 · 13 |
cs destroy | 01 · 02 · 03 · 04 · 05 · 11 · 12 · 15 |
Access¶
| Command | Scenarios that run it |
|---|---|
cs ssh | 01 · 02 · 03 · 11 |
cs vpn status | 12 |
cs vpn add-user | 12 |
cs vpn revoke | 12 |
cs vpn users | 12 |
cs vpn connect | 12 |
cs vpn disconnect | 12 |
cs vpn provision | 12 |
Kubernetes¶
| Command | Scenarios that run it |
|---|---|
cs k8s info | 03 · 04 · 05 |
cs k8s kubeconfig | 03 · 05 · 16 |
cs k8s tunnel | 03 · 16 |
cs k8s untunnel | 03 · 16 |
cs env show | 05 |
cs env use | 05 |
cs env clear | 05 |
cs node add | 04 · 13 |
cs node list | 03 · 04 · 05 · 13 |
cs node remove | 04 · 13 |
cs node scale | 03 · 04 · 13 |
cs kubectl | 03 · 04 · 05 · 06 · 07 · 08 |
cs helm | 04 · 05 · 06 |
cs k9s | 05 |
Platform catalog¶
| Command | Scenarios that run it |
|---|---|
cs platform list | 06 |
cs platform info | 06 · 07 · 08 |
cs platform plan | 06 · 07 · 08 · 09 · 11 |
cs platform install | 06 · 07 · 08 · 15 |
cs platform uninstall | 06 · 07 · 08 · 09 |
cs platform status | 06 · 07 |
cs platform ui | 06 · 07 |
cs platform template | 06 |
Resilience and security¶
| Command | Scenarios that run it |
|---|---|
cs dr status | 08 · 18 |
cs dr backup | 08 · 18 |
cs dr restore | 08 |
cs dr backups | 08 · 18 |
cs dr schedule | 08 · 17 |
cs dr test | 08 |
cs dr describe | 08 |
cs dr logs | 08 |
cs chaos run | 09 |
cs chaos list | 09 |
cs chaos status | 09 |
cs chaos stop | 09 |
cs chaos report | 09 |
cs scan cis | 10 |
cs scan kube | 10 |
cs scan images | 10 |
cs scan host | 10 |
cs scan stig | 10 |
cs scan cloud | 04 · 10 |
cs scan fips | 10 · 11 |
cs scan architecture | 10 · 16 |
cs scan all | 10 |
cs scan reports | 10 · 11 |
cs evidence list | 14 |
cs evidence read | 14 |
Operational readiness¶
| Command | Scenarios that run it |
|---|---|
cs ops list | 16 |
cs ops credentials-backend | 19 |
cs ops acceptance | 19 |
cs ops release-verify | 19 |
cs ops expiry-plan | 17 |
cs ops expiry-cleanup | 17 |
cs ops health | 16 · 18 |
cs ops network | 16 |
cs ops profile | 17 |
cs ops spec-export | 17 |
cs ops spec-validate | 17 |
cs ops spec-diff | 17 |
cs ops spec-import | 17 |
cs ops policy-check | 17 |
cs ops drift | 18 |
cs ops upgrade-plan | 18 |
cs ops upgrade-apply | 18 |
cs ops recovery-plan | 18 |
cs ops recovery-test | 18 |
Cost and data¶
| Command | Scenarios that run it |
|---|---|
cs finops estimate | 02 · 03 · 04 · 12 · 15 |
cs finops cloud | 15 |
cs finops k8s | 15 |
cs finops report | 15 |
cs databricks connect | 07 |
cs databricks test | 07 |
cs databricks status | 07 |
cs databricks <cli args> | 07 |
cs snowflake connect | 07 |
cs snowflake test | 07 |
cs snowflake status | 07 |
cs snowflake <cli args> | 07 |
AI agents and MCP¶
| Command | Scenarios that run it |
|---|---|
cs enable agentic | 14 |
cs enable headliner | 14 |
cs enable mcp | 14 |
cs enable ui | 15 |
cs disable agentic | 14 |
cs disable headliner | 14 |
cs disable mcp | 14 |
cs disable ui | 15 |
cs agents | 14 |
cs use | 14 |
cs model | 14 |
cs agentic | 14 · 15 |
cs do | 14 |
cs skill list | 14 |
cs skill install | 14 |
cs skill show | 14 |
cs usage report | 14 |
cs usage install | 14 |
cs mcp setup | 14 |
cs mcp status | 14 |
cs mcp guide | 14 |
cs mcp connect | 14 |
cs mcp disconnect | 14 |
cs mcp tools | 14 |
cs mcp config | 14 |
cs mcp test | 14 |
cs mcp serve | 14 |
cs mcp start | 14 |
cs mcp stop | 14 |
cs mcp restart | 14 |
cs mcp logs | 14 |
cs mcp token | 14 |
cs mcp uninstall | 14 |
Console, safety and help¶
| Command | Scenarios that run it |
|---|---|
cs ui open | 15 |
cs ui start | 15 |
cs ui status | 15 |
cs ui stop | 15 |
cs ui restart | 15 |
cs ui logs | 15 |
cs ui token | 15 |
cs ui serve | 15 |
cs undo | 06 · 08 · 13 |
cs undo --list | 13 · 14 |
cs undo --global | 13 |
cs undo --id | 13 |
cs undo --drop | 13 |
cs creds list | 13 |
cs creds set | 02 · 11 · 12 · 13 · 14 |
cs creds unset | 11 · 13 |
cs creds clear | 13 |
cs explain | 01 · 02 · 03 · 04 · 05 · 06 · 07 · 08 · 09 · 10 · 11 · 12 · 13 · 14 · 15 |
cs help | 01 · 02 · 03 · 04 · 05 · 08 · 10 · 11 · 12 · 13 · 14 |
Tooling¶
| Command | Scenarios that run it |
|---|---|
cs install | 03 · 04 · 05 · 10 · 12 · 14 |
cs deps status | 02 |
cs deps install | 02 |
cs deps image | 02 |
cs deps bundle | 02 |
cs deps runtime | 02 |