Prism — Crossdeck

Render an interactive line chart of unique visitors and page views over time for a host you own, and return the period totals as text. Returns a rendered chart in MCP-Apps-capable hosts (a text summary elsewhere) plus the totals. Use when asked to chart, graph, draw, or visualize growth or traffic for a subdomain. The host must be a verified origin of the project, or the request is rejected.

Answer 'what's my biggest signup source?'. Returns new customers grouped by their TRUE first-touch source — the referrer/utm on their earliest anonymous visit (e.g. chatgpt.com, google), carried forward to the customer at signup, NOT the signup-moment referrer. Counts only, highest-first. Forward-only: coverage.state='not_instrumented' means no signups have been attributed since first-touch tracking shipped, NOT zero signups historically. Use for acquisition-channel and signup-source questions.

Cross-match one customer across every layer Crossdeck joins by identity. Returns what they pay (monthly cents), their active entitlement count, and their database read-cost — in one view. Identify the person by ANY ONE of the identifiers below; they all resolve to the same canonical customer. Use for 'how much does this user pay and what do they cost us?'. If no customer resolves, returns a no-match result, not an error.

The canonical count of an app's customers, sourced from the revenue + identity ledgers — use THIS for 'how many users/customers does this app have', NOT get_read_cost (whose attributedActors is a cost artifact, not a user count). Returns payingCustomers (distinct customers with an active subscription — a real point-read), plus totalCustomers and activePeople. Each is state-tagged in data.coverage.<field>.state: payingCustomers is `measured`; totalCustomers and activePeople are `not_instrumented` today (Crossdeck doesn't maintain those counters yet — the response says so, and null there is NOT zero). Read the coverage state before reporting any of them.

For one error (by fingerprint), get the PAYING users it actually hit — the row-level moat answer: which paying customer broke, and how much are they worth. Returns a list sorted by revenue at risk (highest first), each row with your own user id, monthly revenue in cents, and when they last hit the error. Joins the error layer to identity and revenue. Returns only your own identifiers — never emails or names you didn't provide. Use when you need the actual customers, not just the counts from get_error_impact.

For one error (by fingerprint/issue id), get its blast radius joined to identity. Returns distinct users affected, how many of those are PAYING customers, and the error's type, status, occurrence count, and first/last-seen timestamps — counts only. Use to size an error's impact. For the NAMED paying users behind the counts (and their revenue at risk), call get_error_affected_users instead.

Get headless web analytics for one host/subdomain the app owns (e.g. a tenant's subdomain). Returns page views and unique visitors with totals; granularity='day' adds a daily series. The host MUST be a verified origin of the project, or the request is rejected (403). Use for per-tenant or per-subdomain traffic questions.

Get the top pages or top referrers for one host the app owns, ranked and paginated. Returns rows of { key, views } highest-first, where `key` is a page path (default) or a referrer. Set dimension='top_referrers' for traffic sources. The host must belong to the selected app. Use for 'most-viewed pages on this subdomain' or 'where its traffic comes from'.

The coverage manifest — LOAD THIS FIRST (workspace key). For every app it returns which surfaces are instrumented vs a blind spot: coverage.revenue / readCost / analytics / userCensus, each 'measured' or 'not_instrumented'. Use it to interpret every later number: if an app's readCost is 'not_instrumented' here, a blank read-cost for it later is a KNOWN blind spot, not a real zero. Prevents reporting an un-wired surface as fact.

Get an app's database read-cost over the last `days` days: total metered reads split into per-user reads vs un-attributed overhead, plus the breakdown by operation. `attributedActors` counts the distinct actor keys the SDK labelled during cost reporting — it INCLUDES background/service actors and is NOT a user or customer count (for that, call get_customers or get_revenue). Server-side reads only (excludes the outbound API + iOS). When an app has no read-cost attribution the fields come back `not_instrumented` (a blind spot), not 0 — check data.coverage.state. Use for 'what's driving our database reads' or 'which operation costs the most'.

Get an app's recurring revenue from its maintained revenue ledger (a point-read, never a scan). Returns MRR in cents, the paying-customer count, and the per-rail split across Stripe, Apple, and Google. With granularity='day' it adds a daily MRR time series (`data.series`, oldest→newest) — pass days to set the window (e.g. days=56 for 8 weeks) and use it to answer 'is my MRR up or down over time', 'MRR history', or any trend question. Use also for 'what's our MRR' or 'how many paying customers'. An app with no payment rail connected returns `not_instrumented` (check data.coverage.state) — that is NOT $0 earned. For a broader customer census (total / active people, not just paying), call get_customers.

Get an app's revenue AVERAGES from maintained ledgers (point-reads, no scan): ARPU (avg MRR per paying customer) and average lifetime value per paying customer. Lifetime value = combinedLifetimeValueCents (all customers' LTV: active-subscription annualized run-rate + banked one-offs) — it RECONCILES to the dashboard's per-customer LTV and is NOT the same as recognisedCashToDateCents (actual cash banked), which is returned separately. Answers 'average lifetime value per user' and 'what's my ARPU'. Averages are ACROSS THE PAYING BASE — NOT any one customer's pay (use get_customer for one). 'user' = paying customer here; a per-all-users average is not_instrumented. `not_instrumented` when no revenue rail is connected — not $0.

Pivot the whole customer book by an axis Crossdeck owns and rank it — the cross-layer question no single tool answers: 'which paying customers cost me the most in database reads?'. by='read_cost' ranks by metered reads over the window; by='mrr' ranks by monthly revenue. Filter with segment='paying'|'non_paying'|'all'. Each row carries your own user id, monthly revenue (cents), paying flag, and read-cost — joined by identity. Rows are your own identifiers only, never emails/names. Read-cost that isn't instrumented comes back as coverage.state='not_instrumented' (a blind spot), never a misleading empty ranking. Use for 'who are my most expensive paying users' or 'top customers by revenue'.

List every app (project) in your Crossdeck portfolio that this connector can read — the same set your Pulse dashboard shows. Returns an array of { id, name } per app. Call this FIRST when a workspace key (cd_wk_) is connected, then pass an id to other tools or set a default with use_project. With a single-app key it returns just that one app. Read-only; takes no arguments.

Render a customer's cross-layer dashboard — what they pay (monthly), their active entitlements, and their database read-cost, joined by identity — and return the same as a text summary. Returns a rendered dashboard in MCP-Apps-capable hosts (a text summary elsewhere). Identify the customer by ANY ONE of the identifiers below; they all resolve to the same canonical customer. Use when asked to show, open, or visualize a customer's full picture. Returns a no-match result if no customer resolves.

Set the current app for this conversation so later tools default to it without repeating `project`. Behaves as session state: it persists until you call this again and applies to every subsequent tool call. Returns { currentProject } — the project id now in effect. Use when the user says 'switch to <app>' or names an app to focus on; call list_projects first for valid ids. No effect with a single-app key (the project is fixed).