get_recommended_metric_alarms

// WHEN AI AGENTS USE THIS TOOL

AI agents call get_recommended_metric_alarms to retrieve information from AWS Labs CloudWatch MCP Server without modifying any data. This is common in research, monitoring, and reporting workflows where the agent needs context before taking action. Because read operations don't change state, they are generally safe to allow without restrictions -- but you may still want rate limits to control API costs.

// WHY ENFORCE A POLICY ON GET_RECOMMENDED_METRIC_ALARMS

Even though get_recommended_metric_alarms only reads data, uncontrolled read access can leak sensitive information or rack up API costs. An agent caught in a retry loop could make thousands of calls per minute. A rate limit gives you a safety net without blocking legitimate use.

// RECOMMENDED POLICY

Read-only tools are safe to allow by default. No rate limit needed unless you want to control costs.

aws-labs-cloudwatch-mcp-server.yaml

tools:
  get_recommended_metric_alarms:
    rules:
      - action: allow

See the full AWS Labs CloudWatch MCP Server policy for all 19 tools.

// DETAILS

Tool Name get_recommended_metric_alarms

Category Read

MCP Server AWS Labs CloudWatch MCP Server MCP Server

Risk Level Low

// MORE AWS LABS CLOUDWATCH MCP SERVER TOOLS

Destructive cancel_logs_insight_query Execute execute_cwl_insights_batch Execute execute_log_insights_query Execute execute_promql_query Execute execute_promql_range_query Read analyze_log_group Read analyze_metric Read describe_log_groups

View all 19 tools →

// WHAT CAN GO WRONG

Agents calling read-class tools like get_recommended_metric_alarms have been implicated in these attack patterns. Read the full case and prevention policy for each:

Browse the full MCP Attack Database →

// OTHER READ TOOLS ACROSS MCP SERVERS

Other tools in the Read risk category across the catalogue. The same policy patterns (rate-limit, allow) apply to each.

Firecrawl Web Scraping Server firecrawl_check_crawl_status Firecrawl Web Scraping Server firecrawl_crawl Firecrawl Web Scraping Server firecrawl_deep_research Firecrawl Web Scraping Server firecrawl_extract Firecrawl Web Scraping Server firecrawl_map Firecrawl Web Scraping Server firecrawl_scrape

// RELATED READING

// FAQ

What does the get_recommended_metric_alarms tool do? +

Gets recommended alarms for a CloudWatch metric. This tool retrieves alarm recommendations for a specific CloudWatch metric identified by its namespace, metric name, and dimensions. The recommendations are filtered to match the provided dimensions. Usage: Use this tool to get recommended alarm configurations for CloudWatch metrics, including thresholds, evaluation periods, and other alarm settings. Args: ctx: The MCP context object for error handling and logging. namespace: The metric namespace (e.g., "AWS/EC2", "AWS/Lambda") metric_name: The name of the metric (e.g., "CPUUtilization", "Duration") dimensions: List of dimensions with name and value pairs region: AWS region to query. Defaults to AWS_REGION environment variable or us-east-1 if not set. profile_name: AWS CLI Profile Name to use for AWS access. Falls back to AWS_PROFILE environment variable if not specified, or uses default AWS credential chain. statistic: The statistic to use for alarm recommendations. Must match the metric's data type: - Aggregate count metrics (RequestCount, Errors, Faults, Throttles, CacheHits, Connections, EventsProcessed): Use 'Sum' - Event occurrence metrics (Invocations, CacheMisses): Use 'SampleCount' - Utilization metrics (CPUUtilization, MemoryUtilization, DiskUtilization, NetworkUtilization): Use 'Average' - Latency/Time metrics (Duration, Latency, ResponseTime, ProcessingTime, Delay, ExecutionTime, WaitTime): Use 'Average' - Size metrics (PayloadSize, MessageSize, RequestSize, BodySize): Use 'Average' If uncertain about the correct statistic for a custom metric, ask the user to confirm the metric type before generating recommendations. Using the wrong statistic (e.g., 'Average' on Invocations) will produce ineffective alarm thresholds Returns: AlarmRecommendationResult: A result containing alarm recommendations and optional message. Empty recommendations list if no recommendations are found. Example: recommendations = await get_recommended_metric_alarms( ctx, namespace="AWS/EC2", metric_name="StatusCheckFailed_Instance", dimensions=[ Dimension(name="InstanceId", value="i-1234567890abcdef0") ] ) for alarm in recommendations: print(f"Alarm: {alarm.alarmDescription}") print(f"Threshold: {alarm.threshold.staticValue}"). It is categorised as a Read tool in the AWS Labs CloudWatch MCP Server MCP Server, which means it retrieves data without modifying state.

How do I enforce a policy on get_recommended_metric_alarms? +

Add a rule in your Intercept YAML policy under the tools section for get_recommended_metric_alarms. You can allow, deny, rate-limit, or validate arguments. Then run Intercept as a proxy in front of the AWS Labs CloudWatch MCP Server MCP server.

What risk level is get_recommended_metric_alarms? +

get_recommended_metric_alarms is a Read tool with low risk. Read-only tools are generally safe to allow by default.

Can I rate-limit get_recommended_metric_alarms? +

Yes. Add a rate_limit block to the get_recommended_metric_alarms rule in your Intercept policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.

How do I block get_recommended_metric_alarms completely? +

Set action: deny in the Intercept policy for get_recommended_metric_alarms. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.

What MCP server provides get_recommended_metric_alarms? +

get_recommended_metric_alarms is provided by the AWS Labs CloudWatch MCP Server MCP server (awslabs.cloudwatch-mcp-server). Intercept sits as a proxy in front of this server to enforce policies before tool calls reach the server.

get_recommended_metric_alarms

Let agents act without letting them run wild.