Tools Reference

The DigiUsher MCP Server has 45 read-only tools in 12 domains. Every tool needs an authentication, and most tools need an organization_id. To find your organizations, call list_organizations first.

Every tool calls the DigiUsher API for you, so the permissions of your DigiUsher account apply. A tool with a body parameter takes a JSON object. A separate Body fields table gives the fields of that object.

Organizations & Users

list_organizations

Entry-point tool for organization-scoped workflows. Returns organizations the authenticated user can access. Call this first when organization_id is unknown.

Example prompt: "What organizations do I have access to in DigiUsher?"

---

get_current_user

Returns the profile of the authenticated DigiUsher user. Use it to identify the caller, and to make sure that the account context is correct, before an operation on an organization.

Example prompt: "Who am I in DigiUsher?"

---

Data Sources

list_data_sources

Lists cloud account and billing data sources for an organization. Useful for mapping data_source_id values in expense and recommendation results.

Example prompt: "What cloud accounts are connected to my organization?"

---

get_data_source

Get detailed information for a single data source by ID, including configuration and status. Use list_data_sources first to find the data_source_id.

Example prompt: "Show me the details and status of data source abc-123."

---

Expenses

DigiUsher recommends this workflow for the expense tools:

1. Call get_expense_dimension_catalog to find the available dimensions.
2. Call query_expense_dimension_values to get the valid filter values of a dimension.
3. Call query_expense_data to run the expense query.
4. Call get_expense_dimension_lookups to translate the IDs in the results into names.

get_expense_dimension_catalog

Lists available expense dimensions and their capabilities for filtering or grouping.

Example prompt: "What dimensions can I use to break down my cloud costs?"

---

query_expense_dimension_values

Returns the valid values of one expense dimension. It works for standard dimensions and for tags. In a query_expense_data filter, use only these values. Another value gives an error.

Body fields:

The dimension object takes one of four shapes:

Standard dimension names: service_name, service_category, service_subcategory, resource_type, resource_id, resource_group, region_id, network_from, network_to, data_source_id, provider, publisher, charge_category, pricing_category, license_model, sku_name, invoice_id, invoice_issuer, commitment_discount_type, commitment_discount_category, commitment_discount_id, commitment_discount_status, consumed_unit, pool_id, rule_id, allocated_resource_id, allocated_resource_name, allocated_method_id.

Example prompt: "What services are available in my expense data?"

---

query_expense_data

Query aggregated expense data with grouping, filtering, and sorting.

Body fields:

group_by entries: a plain string from the list of standard dimension names, or one of these objects:

filters object: one key per dimension you want to filter. Unknown keys are rejected.

Ordering rules: a cost metric used in order_by must also appear in metrics, and date cannot be used when granularity is total.

Example prompt: "Show me our AWS spending for January 2026 broken down by service, sorted by effective cost descending, top 10."

---

get_expense_dimension_lookups

Returns lookup tables to translate IDs (data_source_id, pool_id, rule_id) into user-friendly names. Use after query_expense_data when results contain IDs.

Example prompt: "Translate the data source IDs in my expense results to names."

---

Anomaly Detection

If you give no start_date and no end_date, the summary tools and the list tools use the last 90 days. A range that you give cannot be longer than 366 days.

get_anomaly_summary

Get aggregated anomaly counts by severity, anomaly type, and group. Use as a dashboard-style overview before drilling into individual anomalies.

Example prompt: "Give me a summary of cost anomalies detected this month."

---

list_anomalies

List anomalies with filtering, sorting, and pagination. Use after get_anomaly_summary to inspect matching anomalies in detail.

Body fields:

Example prompt: "Show me all high-severity cost anomalies from the last 30 days, biggest cost impact first."

---

get_anomaly

Get a single anomaly by ID with computed impact metrics and metadata. Use after list_anomalies for deeper inspection.

Example prompt: "Show me the full details of anomaly abc-123."

---

Recommendations

list_all_scenarios

List recommendation scenarios and their configurable thresholds. Use to discover scenario IDs and threshold metadata before working with overrides or filtering recommendations.

Example prompt: "What types of cost optimization scenarios does DigiUsher check for on AWS?"

---

get_scenario_by_id

Get one recommendation scenario by ID, including threshold definitions and optimization detection metadata.

Example prompt: "Show me the details of the rightsizing scenario."

---

list_threshold_overrides

List threshold overrides configured for the organization, optionally filtered by scenario. Use to audit customized scenario sensitivity.

Example prompt: "Are there any custom threshold overrides configured for our recommendations?"

---

get_threshold_override

Get a specific threshold override by ID. Use after list_threshold_overrides when details for one override are needed.

Example prompt: "Show me override abc-123."

---

list_recommendations

List cost optimization recommendations with filtering, sorting, and pagination. Common filters include status, severity, scenario tags, cloud account, and resource attributes.

Body fields:

Example prompt: "Show me the top 5 open recommendations sorted by annual savings."

---

get_savings_summary

Get aggregated potential savings from recommendations, with optional filtering and grouping. Useful for executive summaries and trend breakdowns.

Body fields: the same filter fields as list_recommendations, without the pagination fields and the sorting fields. These fields are status (default ["open"]), severity, scenario_id, scenario_tags, data_source_id, assignee_id, commitment_id, resource_id, resource_type, region, service, resource_tags, created_at_start_date, created_at_end_date, and group_by.

Example prompt: "What's the total potential savings from all open recommendations, grouped by scenario?"

---

get_recommendation

Get one recommendation by ID, including the latest automation status and detailed payload.

Example prompt: "Show me the full details of recommendation abc-123."

---

KPIs

list_kpi_definitions

List available KPI definitions, including units, categories, and favorable direction. Use before snapshot or time-series calls to select kpi_ids.

Available KPI IDs: effective_savings_rate, compute_commitment_coverage, commitment_discount_waste, allocation_accuracy_index, compute_cost_per_core, cost_per_gb_stored, legacy_resource_percentage, cost_optimization_index.

Example prompt: "What FinOps KPIs are available for my organization?"

---

get_kpi_snapshots

Get KPI values for a single date.

Example prompt: "What are our current KPI values?"

---

get_kpi_time_series

Get KPI values across a date range. start_date must be on or before end_date.

Example prompt: "How has our effective savings rate trended over the last 3 months?"

---

Unit Economics

Unit economics combine a cost expression with a North Star Metric (NSM) to compute cost per unit of business value. Dimension map groups translate raw billing dimension values into business categories that formulas can use.

list_dimension_map_groups

List dimension map groups in the organization. A dimension map group is a named set of mappings used to translate raw billing dimension values into business-meaningful categories.

Example prompt: "What dimension map groups are defined for unit economics?"

---

get_dimension_map_group

Get a single dimension map group by ID, including all of its source-to-target mapping entries.

Example prompt: "Show me the mappings in dimension map group abc-123."

---

list_formulas

List unit-economics formulas in the organization. Use to discover formula IDs before fetching or calculating one.

Example prompt: "What unit economics formulas do we have?"

---

get_formula

Get a single unit-economics formula by ID, including its cost expression, NSM reference, default grouping, and metadata.

Example prompt: "Show me how the cost-per-transaction formula is defined."

---

calculate_formula

Calculate cost, NSM, and cost-per-unit for a formula over a date range. Returns one entry per period with totals plus a per-group breakdown.

Body fields:

Example prompt: "What was our cost per transaction each month last quarter?"

---

Chargeback

get_chargeback_for_month

Get chargeback data for a specific month. Returns persisted runs when available, or on-the-fly computed results for missing periods.

Example prompt: "Show me the chargeback data for January 2026."

---

get_cost_allocation_summary

Get the showback summaries or the chargeback summaries of one or more months. DigiUsher moves start_date to the start of its month. end_date is inclusive, and it can be any date.

Body fields:

Example prompt: "Show me the chargeback allocation across teams for Q1 2026, by month."

---

get_cost_allocation_dimensions

Get available filter dimension values (services, regions, data sources) for cost allocation queries. Use this to discover valid filter options before calling get_cost_allocation_summary.

Example prompt: "What services and regions are available for chargeback filtering?"

---

Forecasting

group_type accepts ORGANIZATION or SHOWBACK_POOL.

get_forecast_summary

Get a forecast overview for the organization or a specific showback pool. Returns projected totals, trend direction, and confidence range. Omit group_type and group_key for the org-level forecast.

Example prompt: "What's the cost forecast for my organization?"

---

get_forecast_series

Get daily forecast time-series data for charting: actuals, predicted values, and confidence bands. Use after get_forecast_summary to drill into the trend.

Example prompt: "Show me the daily forecast trend with confidence bands for next month."

---

get_forecast_batch

Get the forecast summary of every entry of one group_type, in one call. With SHOWBACK_POOL, you get the forecast summary of every pool. ORGANIZATION does not work here. For the organization, use get_forecast_summary.

Example prompt: "Give me forecast summaries for all cost pools at once."

---

Resource Inventory

get_resource_types

List available resource types for resource inventory queries. Use before get_dimension_values to select a valid resource_type.

Example prompt: "What resource types are in my cloud inventory?"

---

get_dimension_values

Get values for resource inventory dimensions for a specific resource type. This is for resource dimensions, not expense dimensions.

Example prompt: "What regions do our EC2 instances run in?"

---

get_resource_time_series

Get the metric time series of one discovered cloud resource. Returns the daily metric values of the date range. The default range is the last 30 days.

Example prompt: "Show me the CPU utilization metrics for resource abc-123 over the last 30 days."

---

Governance

list_tagging_policies

List all tagging policies configured for the organization. Each policy defines required tags, allowed values, and the resources it applies to. Use to discover policy IDs before fetching details or violations.

Example prompt: "What tagging policies are enforced in our organization?"

---

get_tagging_policy

Get a single tagging policy by ID, including its rules, required tag keys, allowed values, and resource scope.

Example prompt: "Show me the rules of the cost-center tagging policy."

---

list_tagging_violations

List individual tagging-policy violations across the organization, optionally filtered to one or more policies and paginated.

Example prompt: "List the 20 most recent resources violating our tagging policies."

---

get_tagging_violations_summary

Get a roll-up summary of tagging violations across the organization, including counts by policy and severity. Use for governance dashboards before drilling into individual violations.

Example prompt: "How many tagging violations do we have, by policy?"

---

AI Attribution

These tools report the spend of the AI coding agents. They need an AI telemetry data source, of the type ai_telemetry. If a tool returns a 404, call list_data_sources and look for such a source. The amounts use the currency of your organization, which is not always USD.

get_ai_spend_overview

Get header KPIs for AI coding-agent spend over one date window: total and estimated spend, spend by model family, blended rate per million tokens, cache hit ratio, run and active-user counts, and cost per active hour. Start here, then use query_ai_activity or list_ai_runs to break the window down.

Example prompt: "How much did we spend on AI coding agents in August, and what was our cache hit ratio?"

---

query_ai_activity

Get a series of AI spend and tokens in buckets, with one breakdown dimension. The tool dimension works in another way. It fills tool_activity with the number of calls and the latency of each tool, and it returns an empty series and empty totals. A tool call has no spend of its own.

Body fields:

Example prompt: "Break down last month's AI agent spend by team, week by week."

---

list_ai_runs

List agent runs (sessions) in the window with their spend, token, and event aggregates. Filter to one user or model, or pass run_id to drill into a single run, which also fills detail with that run's per-model split.

Body fields:

Example prompt: "Which agent sessions cost the most last week?"

---

get_ai_work_items_summary

Get the unit economics of the work items, which are the pull requests, that a person merged in the window. The result gives the cost for one merged work item, the cost for 100 added lines, and the part of the spend of the window that belongs to a work item. merged_items gives each item of the window the full spend of its runs, over the whole life of those runs. window_spend counts only the spend inside the window. You cannot calculate one of the two values from the other value.

Example prompt: "What did each merged pull request cost us in AI spend last month?"

---

list_ai_work_items

List the work items (pull requests) merged in the window with the AI spend attributed to each, the lines they added and deleted, and the quality of the join behind the attribution. Use list_ai_work_item_groups to roll these up by repository or author instead.

Body fields:

Example prompt: "Show me the pull requests merged in the acme/widgets repo last month with the most AI spend attributed."

---

list_ai_work_item_groups

Roll the window's merged work items up by repository, author, or both, with cost per merged work item and cost per 100 added lines for each group. Rollups aggregate the whole window before they paginate, so group totals are complete.

Body fields:

Example prompt: "Compare AI cost per merged pull request across our repositories."