alibabacloud-cms-alert-rule-create

Alibaba Cloud Alert Rule Creation

This skill creates CMS 1.0 alert rules for cloud resource monitoring using CloudMonitor metrics. Supports all CMS-integrated cloud products through dynamic metric discovery.

Workflow

Pre-flight Checklist (MANDATORY)

Before creating ANY alert, complete these API calls to ensure correct workflow execution.

These API calls are required for every alert creation. Always query contacts via the designated tools, even if the values seem known.

Critical Rules

1. Contact Query Before Create (MANDATORY)

This step is REQUIRED and CANNOT be skipped.

1. MUST call describe-contact-group-list before creating any CMS alert
2. User provided contact name → Fuzzy match against existing groups
3. If no match → Help user create new contact group

2. Resources Parameter (MANDATORY)

The --resources parameter MUST always be explicitly passed. Never omit this parameter.

• All resources: --resources '[{"resource":"_ALL"}]'
• Specific instances: --resources '[{"resource":"i-xxx"}]' or --resources '[{"resource":"i-xxx"},{"resource":"i-yyy"}]'

This applies to ALL products (ECS, RDS, SLB, OSS, MongoDB, etc.).

3. Required API Calls Summary

4. Contact Group Fuzzy Matching

When user mentions a contact group, apply these matching rules:

5. CLI Command Timeout

All aliyun CLI commands MUST be executed with a timeout to prevent hanging:

• Default timeout: 30 seconds for query operations (describe/list/get)
• Extended timeout: 60 seconds for write operations (put/create/update)
• If a command does not return within the timeout, retry once before reporting failure.

6. Duplicate Alert Pre-check

Before creating an alert rule, check if a rule with the same configuration already exists:

• Call describe-metric-rule-list --namespace <ns> --metric-name <metric> and check for matching rules
• If a duplicate exists, inform the user and ask whether to skip or create with a new name

7. Network Access Restriction

This skill only accesses Alibaba Cloud OpenAPI endpoints. Allowed domains:

• cms.aliyuncs.com — CloudMonitor API
• ecs.aliyuncs.com — ECS instance query
• rds.aliyuncs.com — RDS instance query
• slb.aliyuncs.com — SLB instance query

No other external network access is required or permitted.

8. Dynamic Metric Discovery

MUST call describe-metric-meta-list API to discover metrics for the target namespace. DO NOT hardcode metric names. Use metrics.md only as fallback when API is unavailable.

1. Call describe-project-meta to list all available namespaces (when product is unclear)
2. Call describe-metric-meta-list --namespace <ns> to get available metrics
3. Match returned metrics to user's intent (CPU, memory, disk, network, etc.)
4. Fall back to metrics.md only when API call fails

9. CLI Self-Discovery

When unsure about CLI command syntax, arguments, or available subcommands, use --help to discover:

# List all available CMS commands
aliyun cms --help
 
# Show detailed usage for a specific command
aliyun cms <command> --help
# Example:

This is the preferred way to resolve CLI uncertainties rather than guessing parameters.

10. Mandatory Confirmation Before Execution

MUST present configuration summary and get explicit user confirmation BEFORE calling PutResourceMetricRule.

Even if all parameters are clear, DO NOT execute directly. Always show a configuration summary including: Product, Metric, Threshold, Severity, Resources scope, and Contact Group. Wait for user's explicit "Yes" or confirmation before proceeding.

Severity Levels

Reference Files

Prerequisites

Important: All aliyun CLI calls in this skill MUST include the User-Agent header. Set the environment variable before executing any commands.