mcp/billing-cost-management-mcp-server repository overview

⁠AWS Billing and Cost Management MCP Server

Billing and cost management.

What is an MCP Server?⁠

⁠MCP Info

⁠Image Building Info

⁠Available Tools (33)

---

⁠Tools Details

⁠Tool: aws-pricing

Comprehensive AWS pricing analysis tool that provides access to AWS service pricing information and cost analysis capabilities.

This tool supports four main operations:

1. get_service_codes: Get a comprehensive list of AWS service codes from the AWS Price List API
2. get_service_attributes: Get filterable attributes for a specific AWS service's pricing
3. get_attribute_values: Get all valid values for a specific attribute of an AWS service
4. get_pricing_from_api: Get detailed pricing information from AWS Price List API with optional filters

USE THE OPERATIONS IN THIS ORDER:

1. get_service_codes: Entry point - discover available AWS services and their unique service codes. Note that service codes may not match your expectations, so it's best to get service codes first.
2. get_service_attributes: Second step - understand which dimensions affect pricing for a chosen service
3. get_attribute_values: Third step - get possible values you can use in pricing filters
4. get_pricing_from_api: Final step - retrieve actual pricing data based on service and filters If you deviate from this order of operations, you will struggle to form the correct filters, and you will not get results from the API

IMPORTANT GUIDELINES:

• When retrieving foundation model pricing, always use the latest models for comparison
• For database compatibility with services, only include confirmed supported databases
• Providing less information is better than giving incorrect information
• Price list APIs can return large data volumes. Use narrower filters to retrieve less data when possible
• Service codes often differ from AWS console names (e.g., 'AmazonES' for OpenSearch) Parameters|Type|Description -|-|- operation|string|The pricing operation to perform ('get_service_codes', 'get_service_attributes', 'get_attribute_values', 'get_pricing_from_api') attribute_name|string optional|Attribute name (e.g., 'instanceType', 'location', 'storageClass'). Required for get_attribute_values operation. filters|string optional|Optional filters for pricing queries as a JSON string. Format: '{"instanceType": "t3.medium", "location": "US East (N. Virginia)"}' max_results|string optional|Maximum number of results to return (optional) service_code|string optional|AWS service code (e.g., 'AmazonEC2', 'AmazonS3', 'AmazonES'). Required for get_service_attributes, get_attribute_values, and get_pricing_from_api operations.

---

⁠Tool: bcm-pricing-calc

Allows working with workload estimates using the AWS Billing and Cost Management Pricing Calculator API.

IMPORTANT USAGE GUIDELINES:

• Always first check the rate preference setting for the authorized principal by calling the get_preferences operation.
• DO NOT state assumptions about Free Tier API

USE THIS TOOL FOR:

• Listing available workload estimates for the logged in account.
• Filter list of available workload estimates using name, status, created date, or expiration date.
• Get details of a workload estimate.
• Get the list of services, usage type, operation, and usage amount modeled within a workload estimate.
• Get rate preferences set for Pricing Calculator. These rate preferences denote what rate preferences can be used by each account type in your organization.

⁠OPERATIONS

1. list_workload_estimates - list of available workload estimates Required: operation="list_workload_estimates" Optional: created_after, created_before, expires_after, expires_before, status_filter, name_filter, name_match_option, next_token, max_results Returns: List of all workload estimates for the account.

2. get_workload_estimate - get details of a workload estimate Required: operation="get_workload_estimate", identifier Returns: Details of a specific workload estimate.

3. list_workload_estimate_usage - list of modeled usage lines within a workload estimate Required: operation="get_workload_estimate", identifier Optional: usage_account_id_filter, service_code_filter, usage_type_filter, operation_filter, location_filter, usage_group_filter, next_token, max_results Returns: List of usage associated with a workload estimate.

4. get_preferences - get the rate preferences available to an account Required: operation="get_preferences" Returns: Retrieves the current preferences for AWS Billing and Cost Management Pricing Calculator. Parameters|Type|Description -|-|- operation|string| created_after|string optional| created_before|string optional| expires_after|string optional| expires_before|string optional| identifier|string optional| location_filter|string optional| max_pages|string optional| max_results|string optional| name_filter|string optional| name_match_option|string optional| next_token|string optional| operation_filter|string optional| service_code_filter|string optional| status_filter|string optional| usage_account_id_filter|string optional| usage_group_filter|string optional| usage_type_filter|string optional|

---

⁠Tool: budgets

Retrieves AWS budget information using the AWS Budgets API.

This tool uses the DescribeBudgets API to retrieve all budgets for an account.

The API returns information about:

• Budget names, types, and time periods
• Budget limits (amount and unit)
• Current actual spend
• Forecasted spend
• Cost filters applied to budgets

With this information, you can determine which budgets have been exceeded or are projected to exceed their limits.

The tool automatically retrieves the AWS account ID of the calling identity or uses the provided account_id.

---

⁠Tool: compute-optimizer

Retrieves recommendations from AWS Compute Optimizer.

IMPORTANT USAGE GUIDELINES:

• Focus on recommendations with the highest estimated savings first
• Include all relevant details when presenting specific recommendations

USE THIS TOOL FOR:

• Performance optimization (CPU, memory, network utilization analysis)
• Performance-based rightsizing (not cost-based)

DO NOT USE FOR: Cost optimization or idle detection (use cost-optimization-hub)

Note: Compute Optimizer is a regional service. Specify a region to get recommendations for resources in that region. If omitted, defaults to AWS_REGION env var or us-east-1.

This tool supports the following operations:

1. get_ec2_instance_recommendations: Get recommendations for EC2 instances
2. get_auto_scaling_group_recommendations: Get recommendations for Auto Scaling groups
3. get_ebs_volume_recommendations: Get recommendations for EBS volumes
4. get_lambda_function_recommendations: Get recommendations for Lambda functions
5. get_rds_recommendations: Get recommendations for RDS instances
6. get_ecs_service_recommendations: Get recommendations for ECS services

Each operation can be filtered by AWS account IDs, regions, finding types, and more.

Common finding types include:

• UNDERPROVISIONED: The resource doesn't have enough capacity
• OVERPROVISIONED: The resource has excess capacity and could be downsized
• OPTIMIZED: The resource is already optimized
• NOT_OPTIMIZED: The resource can be optimized but specific finding type isn't available Parameters|Type|Description -|-|- operation|string|The operation to perform (e.g., 'get_ec2_instance_recommendations') account_ids|string optional|Optional list of AWS account IDs as JSON array string filters|string optional|Optional filter expression as JSON string max_results|string optional|Maximum number of results to return (1-100) next_token|string optional|Optional pagination token from a previous response region|string optional|AWS region to query (e.g., 'us-west-2'). Defaults to AWS_REGION env var or us-east-1.

---

⁠Tool: cost-anomaly

Retrieves AWS cost anomalies using the Cost Explorer GetAnomalies API.

This tool allows you to retrieve cost anomalies detected on your AWS account during a specified time period. Anomalies are available for up to 90 days.

You can filter anomalies by:

• Date range (required)
• Monitor ARN (optional)
• Feedback status (optional)
• Total impact (optional)

Note: Both start_date and end_date are INCLUSIVE. To get anomalies including today, use today's date as end_date.

Feedback status options:

• YES: Anomalies marked as accurate
• NO: Anomalies marked as inaccurate
• PLANNED_ACTIVITY: Anomalies marked as planned activities Parameters|Type|Description -|-|- end_date|string|End date in YYYY-MM-DD format (inclusive — use today's date to include today's anomalies). Required. start_date|string|Start date in YYYY-MM-DD format (inclusive). Required. feedback|string optional|Optional filter for anomalies by feedback status (YES, NO, PLANNED_ACTIVITY). max_results|string optional|Optional maximum number of results to return. monitor_arn|string optional|Optional ARN of a specific cost anomaly monitor to filter results. total_impact_end|string optional|Optional end value for total impact filter (required when using BETWEEN operator). total_impact_operator|string optional|Optional numeric operator for filtering by total impact (EQUAL, GREATER_THAN, LESS_THAN, GREATER_THAN_OR_EQUAL, LESS_THAN_OR_EQUAL, BETWEEN). total_impact_start|string optional|Optional start value for total impact filter.

---

⁠Tool: cost-comparison

Retrieves AWS cost comparisons between two one-month periods.

Do not use this tool except for comparing the costs of one month to the costs of another month. This tool should not be used for week-over-week or quarter-over-quarter (e.g., comparing Q2 vs. Q1) analysis.

USE THIS TOOL ONLY FOR:

• Month-to-month cost variance analysis (e.g., January vs February)
• Root cause analysis of cost changes between specific months
• Detailed cost driver identification (what exactly caused the cost change)
• Service-level impact analysis for month-over-month changes
• Executive reporting on monthly cost variances

STRICT LIMITATIONS:

• ONLY compares exactly one month to another month
• Both periods must start on 1st day of month, end on 1st day of next month
• Cannot compare weeks, quarters, or custom periods
• DO NOT USE for general cost analysis or flexible time periods

This tool supports two main operations:

1. getCostAndUsageComparisons: Compare costs between two time periods with flexible grouping and filtering
2. getCostComparisonDrivers: Identify key factors driving cost changes between two time periods

Both operations require:

• BaselineTimePeriod: Earlier time period for comparison (must be exactly one month)
• ComparisonTimePeriod: Later time period for comparison (must be exactly one month)
• MetricForComparison: The cost metric to compare (e.g., BlendedCost, UnblendedCost)

Supported metrics for comparison include:

• AmortizedCost: Costs with upfront and recurring reservation fees spread across the period
• BlendedCost: Average cost of all usage throughout the billing period
• NetAmortizedCost: Amortized cost after discounts
• NetUnblendedCost: Unblended cost after discounts
• NormalizedUsageAmount: Normalized usage amount
• UnblendedCost: Actual costs incurred during the specified period
• UsageQuantity: Usage amounts in their respective units

You can group results by dimensions such as:

• SERVICE: AWS service (e.g., Amazon EC2, Amazon S3)
• LINKED_ACCOUNT: Member accounts in an organization
• REGION: AWS Region
• USAGE_TYPE: Type of usage (e.g., BoxUsage:t2.micro)
• INSTANCE_TYPE: EC2 instance type (e.g., t2.micro, m5.large)
• PLATFORM: Operating system (e.g., Windows, Linux)
• TENANCY: Instance tenancy (e.g., shared, dedicated)
• RECORD_TYPE: Record type (e.g., Usage, Credit, Tax)
• LEGAL_ENTITY_NAME: AWS seller of record

Note:

• Time periods must start and end on the first day of a month, with a duration of exactly one month
• The getCostComparisonDrivers operation automatically includes SERVICE and USAGE_TYPE dimensions
• Data is available for the last 13 months, or up to 38 months if multi-year data is enabled Parameters|Type|Description -|-|- baseline_end_date|string|Baseline period end date in YYYY-MM-DD format (must be first day of next month) baseline_start_date|string|Baseline period start date in YYYY-MM-DD format (must be first day of month) comparison_end_date|string|Comparison period end date in YYYY-MM-DD format (must be first day of next month) comparison_start_date|string|Comparison period start date in YYYY-MM-DD format (must be first day of month) metric_for_comparison|string|The cost metric to compare (e.g., BlendedCost, UnblendedCost, AmortizedCost) operation|string|The operation to perform: 'getCostAndUsageComparisons' or 'getCostComparisonDrivers' billing_view_arn|string optional|Optional ARN of a specific billing view filter|string optional|Optional filter to apply to the results as a JSON string group_by|string optional|Optional grouping of results as a JSON string (e.g., '[{"Type": "DIMENSION", "Key": "SERVICE"}]') max_results|string optional|Maximum number of results to return (default: 10, max: 2000 for comparisons, max: 10 for drivers)

---

⁠Tool: cost-explorer

Retrieves AWS cost and usage data using the Cost Explorer API.

IMPORTANT USAGE GUIDELINES:

• Use UnblendedCost metric by default (not BlendedCost) unless user specifies otherwise
• Exclude record_types 'Credit' and 'Refund' by default unless user requests inclusion
• Choose DAILY granularity for periods <3 months, MONTHLY for longer periods
• Start with high-level dimensions (SERVICE, LINKED_ACCOUNT) before detailed ones
• Always remember that the end_date is exclusive

USE THIS TOOL FOR:

• Historical cost trends and spending analysis (any time period)
• Usage pattern analysis over time
• Cost breakdown by service, account, region, or any dimension
• Forecasting future costs and usage
• Resource-level cost analysis (last 14 days)
• Multi-dimensional cost analysis with complex grouping

⁠OPERATIONS

1. getCostAndUsage — account-level historical cost/usage Required: operation="getCostAndUsage", start_date, end_date, granularity, metrics Optional: group_by, filter, next_token, max_pages, billing_view_arn Example: {"operation": "getCostAndUsage", "start_date": "2024-01-01", "end_date": "2024-02-01", "granularity": "DAILY", "metrics": ["UnblendedCost"], "group_by": "[{"Type": "DIMENSION", "Key": "SERVICE"}]"}

1. getCostAndUsageWithResources - Resource-level cost data (limited to last 14 days) Required: operation="getCostAndUsageWithResources", filter, granularity, start_date, end_date Optional: metrics, group_by, next_token, max_pages, billing_view_arn Notes: RESOURCE_ID must be included in either filter OR group_by parameters. This operation is limited to past 14 days of data from current date. Hourly granularity is only available for EC2-Instances resource-level data. All other resource-level data is available at daily granularity. Example: {"operation": "getCostAndUsageWithResources", "start_date": "2025-08-07", "end_date": "2025-08-21", "granularity": "DAILY", "filter": "{"Dimensions": {"Key": "SERVICE", "Values": ["Amazon Elastic Compute Cloud - Compute"]}}", "group_by": "[{"Type": "DIMENSION", "Key": "RESOURCE_ID"}]"} Returns: Cost data with resource-level granularity

2. getDimensionValues - List of available values for specified dimension Required: operation="getDimensionValues", dimension, start_date, end_date Optional: context, search_string, filter, max_results, billing_view_arn Example: {"operation": "getDimensionValues", "dimension": "SERVICE", "start_date": "2024-01-01", "end_date": "2024-02-01"} Returns: List of values for specified dimension with automatic pagination

3. getCostForecast - Future cost projections Required: operation="getCostForecast", metric, granularity, start_date, end_date Optional: filter, prediction_interval_level, billing_view_arn Example: {"operation": "getCostForecast", "metric": "UNBLENDED_COST", "granularity": "MONTHLY", "start_date": "2025-08-22", "end_date": "2025-11-22"} Notes: metric value for this operation should be in all caps Returns: Cost forecast for specified time period and granularity

4. getUsageForecast - Future usage projections Required: operation="getUsageForecast", metric, granularity, start_date, end_date, filter Optional: prediction_interval_level, billing_view_arn Example 1: {"operation": "getUsageForecast", "metric": "USAGE_QUANTITY", "granularity": "MONTHLY", "start_date": "2025-08-22", "end_date": "2025-11-22", "filter": "{"Dimensions": {"Key": "USAGE_TYPE_GROUP", "Values": ["EC2-Instance"]}}"} Example 2: {"operation": "getUsageForecast", "metric": "USAGE_QUANTITY", "granularity": "MONTHLY", "start_date": "2025-08-22", "end_date": "2025-11-22", "filter": "{"And": [{"Dimensions": {"Key": "SERVICE", "Values": ["Amazon Elastic Compute Cloud - Compute"]}}, {"Dimensions": {"Key": "USAGE_TYPE", "Values": ["BoxUsage:p4de.24xlarge"]}}]}"} Example 3: {"operation": "getUsageForecast", "metric": "USAGE_QUANTITY", "granularity": "MONTHLY", "start_date": "2025-08-22", "end_date": "2025-11-22", "filter": "{"Dimensions": {"Key": "USAGE_TYPE", "Values": ["BoxUsage:p4de.24xlarge", "Reservation:p4de.24xlarge", "UnusedBox:p4de.24xlarge"]}}", "group_by": "[{"Type": "DIMENSION", "Key": "REGION"}]"} Notes: Valid values for metric is: USAGE_QUANTITY, NORMALIZED_USAGE_AMOUNT. Valid values for granularity is: DAILY, MONTHLY. Filter is REQUIRED and must specify USAGE_TYPE or USAGE_TYPE_GROUP to define what usage units to forecast. Returns: Usage forecast for specified time period and granularity

5. getTagsOrValues - Available cost allocation tags or values Required: operation="getTagsOrValues" Optional: start_date, end_date, search_string, next_token, max_pages, billing_view_arn Example 1: {"operation": "getTagsOrValues"} Example 2: {"operation": "getTagsOrValues", "tag_key": "Environment"} Returns: List of available cost allocation tags with automatic pagination. If tag values for a particular key are needed, pass the tag key as a parameter.

6. getCostCategories - Available cost categories or values inside a category Required: operation="getCostCategories", start_date, end_date Optional: cost_category_name, search_string, next_token, max_pages, billing_view_arn Example 1 (list category names): {"operation": "getCostCategories", "start_date": "2024-01-01", "end_date": "2024-08-01"} Example 2 (list values inside a category): {"operation": "getCostCategories", "cost_category_name": "CostCenters", "start_date": "2024-01-01", "end_date": "2024-08-01"} Notes: Without cost_category_name the response contains CostCategoryNames. With cost_category_name set the response contains CostCategoryValues for that category. Pagination is automatic.

7. getSavingsPlansUtilization - Savings Plans utilization data Required: operation="getSavingsPlansUtilization", start_date, end_date Optional: granularity, filter Example: {"operation": "getSavingsPlansUtilization", "granularity": "MONTHLY"} Notes: This operation supports only DAILY and MONTHLY granularity Returns: Savings Plans utilization for the specified time period

DIMENSION REFERENCE:

• AZ: The Availability Zone (e.g., us-east-1a)
• DATABASE_ENGINE: The Amazon RDS database (e.g., Aurora, MySQL)
• DEPLOYMENT_OPTION: RDS deployment scope (SingleAZ, MultiAZ)
• INSTANCE_TYPE: The EC2 instance type (e.g., m4.xlarge)
• INSTANCE_TYPE_FAMILY: Family of instances (e.g., Compute Optimized, Memory Optimized)
• LINKED_ACCOUNT: AWS member accounts
• OPERATING_SYSTEM: OS type (e.g., Windows, Linux)
• PLATFORM: EC2 operating s

[...]