Imported from huaweicloud/huaweicloud-skills (skills/container/ucs/huawei-cloud-ucs-policy-governor/SKILL.md). Install upstream with npx skills add huaweicloud/huaweicloud-skills --skill huawei-cloud-ucs-policy-governor. Copyright stays with the author.

Huawei Cloud UCS Policy Governor

Overview

This skill provides policy governance and compliance management capabilities for Huawei Cloud UCS (Ubiquitous Cloud Native Service) using the hcloud CLI, covering policy instance lifecycle, policy definitions, policy enforcement, and compliance auditing.

Architecture: hcloud CLI → UCS Service API → PolicyInstance/PolicyDefinition/PolicyJob resources

Related Skills:

• huawei-cloud-ucs-cluster-onboarding-manager - Cluster registration, lifecycle, fleet grouping, and access management

Capabilities:

• Create policy instances for clusters or fleet groups
• Update, query, and delete policy instances
• List and query policy definitions (templates)
• Enable and disable policies on clusters or fleet groups
• Check policy enforcement job status via ListPolicyJobs/ShowPolicyJob
• Audit fleet compliance and review policy enforcement results

Typical Use Cases:

• "Create a security policy instance for my production cluster"
• "Create a compliance policy for my fleet group"
• "List all available policy definitions"
• "Enable a policy on cluster 'prod-backend'"
• "Enable a policy on fleet group 'production-fleet'"
• "Disable a policy temporarily for maintenance"
• "Check policy enforcement job status"
• "Audit policy enforcement across all clusters"
• "Update a policy instance configuration"
• "Delete an obsolete policy instance"
• "Query policy definition details before applying"

Workflow

The skill workflow is as follows:

1. Environment Check — Verify hcloud CLI is installed and AK/SK credentials are configured (see CLI Installation Guide)
2. Policy Definition Query — Call ListPolicyDefinitions to browse available governance policy templates
3. Policy Instance Management — Choose management action based on user request:
   • Create: CreateClusterPolicyInstance or CreateClusterGroupPolicyInstance (requires cluster/fleet with policy center enabled)
   • Update: UpdatePolicyInstance (requires user confirmation — see Parameter Confirmation)
   • Query: ListPolicyInstances / ShowPolicyInstance
   • Delete: DeletePolicyInstance (⚠️ irreversible, requires user confirmation)
4. Policy Center Management (optional):
   • Enable: EnableClusterPolicy / EnableClusterGroupPolicy (requires cluster/fleet in Available status)
   • Disable: DisableClusterPolicy / DisableClusterGroupPolicy (requires user confirmation — suspends enforcement)
5. Compliance Audit — Call ListPolicyJobs / ShowPolicyJob to review policy enforcement history and compliance status
6. Verification — Confirm policy instance status and enforcement behavior (see Verification)

KooCLI Command Format Standard

All operations use hcloud UCS <Operation> --<param>=<value> --cli-region=<region> format. Parameters use --key=value style (not JSON body). See Parameter Reference for detailed parameter descriptions.

Supported Regions (verified):

Other regions may be supported but have not been verified. Always test with hcloud UCS ListPolicyDefinitions --cli-region=<region> first.

Prerequisites

1. hcloud CLI Requirements (MANDATORY)

• hcloud CLI installed (version >= 7.2.2)
• Run hcloud version to verify installation
• First-time usage: printf "y\n" | hcloud version to accept privacy statement

2. Credential Configuration

hcloud CLI supports two credential modes via environment variables, automatically detected at runtime:

Mode A — Long-term AK/SK (permanent access):

export HUAWEI_CLOUD_AK=<your-ak>
export HUAWEI_CLOUD_SK=<your-sk>
export HUAWEI_CLOUD_REGION=cn-north-4

Mode B — Temporary AK/SK + SecurityToken (recommended for temporary or delegated access):

export HUAWEI_CLOUD_AK=<your-temp-ak>
export HUAWEI_CLOUD_SK=<your-temp-sk>
export HUAWEI_CLOUD_SECURITY_TOKEN=<your-security-token>
export HUAWEI_CLOUD_REGION=cn-north-4

When HUAWEI_CLOUD_SECURITY_TOKEN is present, hcloud CLI automatically uses temporary credential authentication. When only AK/SK are set, it uses long-term credential authentication.

• Security Rules:
  • 🚫 Never expose AK/SK/SecurityToken values in code, conversation, or commands
  • 🚫 Never use echo $HUAWEI_CLOUD_AK or echo $HUAWEI_CLOUD_SK to check credentials
  • ✅ Use environment variables: HUAWEI_CLOUD_AK, HUAWEI_CLOUD_SK, HUAWEI_CLOUD_REGION, HUAWEI_CLOUD_SECURITY_TOKEN
  • ✅ Prefer IAM users over root account for cloud operations
  • ✅ Enable MFA for sensitive operations

⚠️ Important Security Notes:

• Never commit credentials to version control
• Use IAM users with minimal required permissions
• Enable MFA for sensitive operations
• Rotate AK/SK regularly

3. IAM Permission Requirements

See IAM Permission Policies for complete policy JSON.

Permission Failure Handling:

1. When any command fails due to permission errors, read references/iam-policies.md
2. Display the required permission list and policy JSON to the user
3. Guide the user to create a custom policy in the IAM console and grant authorization
4. Pause execution and wait for user confirmation that permissions have been granted

Core Commands

1. Policy Instance Management

See Task: Policy Management for detailed workflows.

# Create a cluster-level policy instance
hcloud UCS CreateClusterPolicyInstance --clusterid=<ucs-cluster-id> --constraintTemplateID=<template-id> --enforcementAction=deny --namespaces.1=default --namespaces.2=production --parameters='{"maxReplicas":"3"}' --cli-region=cn-north-4

# Create a fleet group-level policy instance
hcloud UCS CreateClusterGroupPolicyInstance --clustergroupid=<fleet-group-id> --constraintTemplateID=<template-id> --enforcementAction=warn --parameters='{"cpuLimit":"2"}' --cli-region=cn-north-4

# Update a policy instance
hcloud UCS UpdatePolicyInstance --policyinstanceid=<instance-id> --constraintTemplateID=<new-template-id> --enforcementAction=warn --parameters='{"cpuLimit":"4"}' --cli-region=cn-north-4

# Show policy instance details
hcloud UCS ShowPolicyInstance --policyinstanceid=<instance-id> --cli-region=cn-north-4

# Delete a policy instance
hcloud UCS DeletePolicyInstance --policyinstanceid=<instance-id> --cli-region=cn-north-4

# List all policy instances (no filter parameters available)
hcloud UCS ListPolicyInstances --cli-region=cn-north-4

2. Policy Definition Management

# List all available policy definitions (no filter parameters available)
hcloud UCS ListPolicyDefinitions --cli-region=cn-north-4

# Show policy definition details
# NOTE: --policydefinitionid requires the metadata.uid (UUID format) from ListPolicyDefinitions,
# NOT the policy definition name. Using a name will return UCS.000000 error.
# Step 1: List policy definitions to get the UID
hcloud UCS ListPolicyDefinitions --cli-region=cn-north-4
# Step 2: Use the metadata.uid value from the list result
hcloud UCS ShowPolicyDefinition --policydefinitionid=<metadata.uid> --cli-region=cn-north-4

3. Policy Enforcement (Enable/Disable)

# Enable a policy on a cluster
hcloud UCS EnableClusterPolicy --clusterid=<ucs-cluster-id> --cli-region=cn-north-4

# Enable a policy on a fleet group
hcloud UCS EnableClusterGroupPolicy --clustergroupid=<fleet-group-id> --cli-region=cn-north-4

# Enable a policy on a cluster with retry
hcloud UCS EnableClusterPolicy --clusterid=<ucs-cluster-id> --retry=true --cli-region=cn-north-4

# Disable a policy on a cluster
hcloud UCS DisableClusterPolicy --clusterid=<ucs-cluster-id> --cli-region=cn-north-4

# Disable a policy on a fleet group
hcloud UCS DisableClusterGroupPolicy --clustergroupid=<fleet-group-id> --cli-region=cn-north-4

4. Policy Enforcement Job Status

See Task: Compliance Audit for detailed workflows.

# List policy enforcement jobs
hcloud UCS ListPolicyJobs --cli-region=cn-north-4

# List policy enforcement jobs filtered by kind
hcloud UCS ListPolicyJobs --kind=EnablePolicy --cli-region=cn-north-4

# Show a specific policy enforcement job
hcloud UCS ShowPolicyJob --jobid=<job-id> --cli-region=cn-north-4

参数确认

Parameter Reference

Common Parameters

Policy Instance Parameters

*Note: --clusterid is required for cluster-level operations (CreateClusterPolicyInstance, EnableClusterPolicy, DisableClusterPolicy). --clustergroupid is required for fleet group-level operations (CreateClusterGroupPolicyInstance, EnableClusterGroupPolicy, DisableClusterGroupPolicy).

Policy Definition Parameters

Policy Job Parameters

Output Format

CreateClusterPolicyInstance / CreateClusterGroupPolicyInstance

[to be verified — UCS responses follow k8s-style format based on verified ShowClusterList/ListPolicyDefinitions patterns]

UCS API returns Kubernetes-style objects, not flat JSON. Based on verified ShowClusterList and ListPolicyDefinitions responses, policy instance responses likely use a k8s-style object structure with kind, apiVersion, metadata, spec, and status fields rather than flat fields like id, constraintTemplateID, enforcementAction.

Key Fields (expected, format to be verified):

• Instance UUID: Likely in metadata.uid (not flat id)
• Constraint template reference: Likely in spec.constraintTemplateID
• Enforcement action: Likely in spec.enforcementAction (warn or deny)
• Status: Likely in status.phase (Enabled, Disabled, Pending)

ListPolicyDefinitions

Response Example (verified):

{
  "items": [
    {
      "kind": "ConstraintTemplate",
      "apiVersion": "templates.gatekeeper.sh/v1beta1",
      "metadata": {
        "name": "k8srequiredresources",
        "uid": "3b900254-0086-11ee-924e-0255ac1000d3",
        "creationTimestamp": "2023-06-01T14:11:41Z",
        "annotations": {
          "name-chinese": "K8sRequiredResources",
          "tag-chinese": "集群安全策略",
          "description-chinese": "..."
        }
      },
      "spec": {
        "type": "general",
        "officialTag": "ClusterSecurityPolicies",
        "level": "1",
        "targetKind": "Pod",
        "official": true,
        "description": "Requires containers to have defined resources set...",
        "constraintTemplate": {
          "kind": "ConstraintTemplate",
          "apiVersion": "templates.gatekeeper.sh/v1",
          "metadata": { "name": "k8srequiredresources" },
          "spec": {
            "crd": {
              "spec": {
                "names": { "kind": "K8sRequiredResources" },
                "validation": { "openAPIV3Schema": { "properties": {} } }
              }
            },
            "targets": [
              {
                "target": "admission.k8s.gatekeeper.sh",
                "rego": "...",
                "libs": []
              }
            ]
          }
        }
      }
    }
  ]
}

Key Fields:

• metadata.name: Constraint template name (used as constraintTemplateID in CreateClusterPolicyInstance, not flat id)
• metadata.uid: Definition UUID
• spec.officialTag: Policy category/tag (not flat category)
• spec.level: Severity level (not flat severity)
• spec.targetKind: Target resource type (e.g., Pod)
• spec.description: Policy description
• spec.constraintTemplate.spec.crd.spec.validation.openAPIV3Schema.properties: Parameter definitions (not flat parameters array)
• spec.type: Policy type (e.g., general)
• spec.official: Whether this is an official (built-in) policy

ListPolicyJobs

Response Example (verified for empty result):

When no jobs exist, returns { "items": [] } (empty array). When populated, returns k8s-style objects based on verified UCS pattern:

{
  "items": []
}

[to be verified for populated response — likely k8s-style objects with kind, apiVersion, metadata, spec, status fields]

Key Fields (expected, format to be verified):

• Job UUID: Likely in metadata.uid (not flat jobid)
• Job type: Likely in spec.kind (EnablePolicy, etc.)
• Job status: Likely in status.phase (Success, Failed, InProgress)

Verification

See Verification Method for step-by-step verification.

Common Region IDs

Best Practices

1. Policy Parameters: Use --constraintTemplateID to reference constraint templates, not --policy_definition_id
2. Fleet-Level Policies: Apply policies to fleet groups using CreateClusterGroupPolicyInstance for consistent enforcement
3. Gradual Rollout: Enable policies on staging clusters first using EnableClusterPolicy, then roll out to production fleet groups using EnableClusterGroupPolicy
4. Compliance Monitoring: Use ListPolicyJobs and ShowPolicyJob to monitor enforcement task status
5. Enforcement Action: Choose warn for initial rollout (violations reported but not blocked), then switch to deny for strict enforcement
6. Disable Before Delete: Disable a policy using DisableClusterPolicy/DisableClusterGroupPolicy before deleting to prevent sudden enforcement gaps
7. Namespace Scoping: Use --namespaces.[N] to scope policy enforcement to specific namespaces

Reference Documents

Notes

• Policy deletion is irreversible — the enforcement configuration is permanently removed
• Disabling a policy suspends enforcement — violations are not checked while the policy is disabled
• Fleet group policies apply to all member clusters — ensure group membership is correct before applying
• AK/SK must never be hardcoded — credentials should only be obtained via environment variables
• hcloud CLI is the only supported method — all operations use hcloud UCS <Operation> format
• CreatePolicyInstance is TWO separate operations — use CreateClusterPolicyInstance for cluster-level and CreateClusterGroupPolicyInstance for fleet group-level policies
• Enable/Disable are scope-specific — use EnableClusterPolicy/DisableClusterPolicy for clusters and EnableClusterGroupPolicy/DisableClusterGroupPolicy for fleet groups
• GetPolicyAssignment does not exist — use ListPolicyJobs and ShowPolicyJob to check enforcement task status
• ListPolicyInstances and ListPolicyDefinitions have no filter parameters — only --cli-region is available
• ShowPolicyDefinition requires UID, not name — the --policydefinitionid parameter must be the metadata.uid (UUID format) from ListPolicyDefinitions response. Using the policy definition name returns UCS.000000 error. Always call ListPolicyDefinitions first to obtain the UID
• hcloud JSON output parsing — the default hcloud output mixes JSON with diagnostic tables, causing json.loads() to fail. For programmatic parsing, use --cli-output=json to get pure JSON output:

# Default output (mixed format, json.loads() will fail)
hcloud UCS ListPolicyDefinitions --cli-region=cn-north-4

# Pure JSON output (safe for json.loads())
hcloud UCS ListPolicyDefinitions --cli-region=cn-north-4 --cli-output=json

Common Pitfalls

See Common Pitfalls & Solutions for detailed troubleshooting guides.

Quick Reference: