Appstack

List saved dashboards, each with the project_id it belongs to. Pass app_id (from whoami) to list the dashboards of the project that app belongs to, or project_id if you already have it. With neither, every project the credential covers is listed — except for a platform-admin credential, which must pick one.

list_dashboards

List the distinct values of a Cube dimension available to your account. Call this before query_metrics to discover valid filter values (e.g. which countries, media sources, or app names exist) instead of guessing — a guessed value silently produces an empty result rather than an error. Args: dimension: A single fully-qualified Cube dimension name, e.g. "events_view.app_name". Call list_metrics first to see what's available to your account. filters: Cube filter objects: {"member", "operator", "values"}. Only for filters other than app_id — that one is added automatically. app_ids: subset of your organization's apps to include; omit to query across all of them. limit: max distinct values to return (default 500).

list_dimension_values

List every integration connected to the caller's apps/projects, with its current status (e.g. "pending", "active", "error"). Call this to check whether a specific integration (Meta Ads, RevenueCat, Superwall, ...) is connected and healthy before asking why its data looks off in query_metrics.

list_integrations

List links, optionally filtered by project_id and/or app_id. Each link includes its full url, redirect links and custom_params (the client-defined query params registered on it, e.g. a promo code). organization_id is optional for a project-scoped credential (defaults to the caller's own organization) but required for a platform-admin credential. A credential covering several organizations lists across all of them unless organization_id narrows it.

list_links

List users (install_id) for an app, each with their most recent event — install_id, event_name, media_source, country, event_ts. assist_network keeps users whose install that ad network assisted without winning its attribution. Call get_user_journey_filters first to discover valid country/ media_source/event_name/assist_network values instead of guessing.

list_user_journeys

Query Appstack Analytics metrics via Cube JS, across your organization's apps. Every member of a single query — measures, dimensions, time_dimension and filters alike — must belong to the SAME view. A query mixing two views is rejected outright, so pick one view from list_metrics and stay inside it. App scoping is enforced automatically (server-side) — do not attempt to filter by app_id yourself, use the app_ids parameter instead. Requesting an app_id outside your organization fails before any query is sent. Call list_dimension_values first if you're unsure what values a filter or dimension actually takes. Args: measures: fully-qualified "<view>.<measure>" names as returned by list_metrics, e.g. ["events_view.total_spend"]. Call list_metrics first to see what's available to your account — a guessed name is rejected, not ignored. dimensions: fully-qualified "<view>.<dimension>" names from the same view as measures, e.g. ["events_view.country"]. filters: Cube filter objects: {"member", "operator", "values"}. Only for filters other than app_id — that one is added automatically. time_dimension: Cube time dimension name, required to use date_range or granularity. date_range: [start, end] ISO dates (requires time_dimension), capped at 60 days inclusive — a wider range is rejected, not clamped, since silently narrowing it could hide data the caller expected. granularity: day|week|month (requires time_dimension). app_ids: subset of your organization's apps to include; omit to query across all of them. limit: max rows to return (default 1000). offset: rows to skip, for paging through results beyond limit. include_total: if true, also return the total number of matching rows (independent of limit/offset). Costs an extra query on the backend, so it's opt-in — leave it off unless you need it. Returns: {"data": [...], "total": <int or null>}. total is only populated when include_total is true.

query_metrics

Report a data inconsistency spotted in a query_metrics result — a negative value where one shouldn't occur, an implausible swing, a zero-count gap, mismatched totals across views, etc. This does not fix or explain the anomaly — it durably records enough context (the exact query that produced it, plus the anomalous rows) for the Appstack team to reproduce and triage, and pings Slack. Call it once you've already decided to warn the user about the anomaly in your response — this is in addition to that warning, not instead of it. Args: anomaly_description: What looks wrong and why, in plain language, e.g. "total_spend is -120 for app_id com.foo.bar on 2026-03-04; ad spend should never be negative." measures: The fully-qualified measures from the query that produced the anomaly, e.g. ["events_view.total_spend"]. dimensions: The dimensions from that query, if any. filters: The filters from that query, if any. time_dimension: The time dimension from that query, if any. date_range: The date_range from that query, if any. observed_values: The specific anomalous row(s) from the result — enough to reproduce, not the full result set.

report_data_anomaly

Update a link's editable metadata (utm_campaign, utm_media_source, experience_type). Only standard (non-ad-network) links can be edited this way.

update_link