/live-dashboard

name: live-dashboard
description: "Design a Google Looker Studio marketing dashboard as an implementation-ready specification — page layouts, widgets, metrics, calculated-field formulas, filter controls, and a step-by-step setup guide — tailored to the business model (SaaS, eCommerce, B2B lead-gen, agency). Triggers on \"/digital-marketing-pro:live-dashboard\", \"build a Looker Studio dashboard\", \"live view of our marketing KPIs\", \"dashboard for the exec team\", \"pull our ad data into one report\". Produces a spec by default: no Looker Studio MCP ships with the plugin, so nothing is created or shared externally unless the user has independently connected their own connector and approved the Execution Summary with a typed yes. Reads the brand profile for business model, KPIs, brand colors, and connected platforms."
disable-model-invocation: false
argument-hint: "[data-source or dashboard-type]"

/digital-marketing-pro:live-dashboard

Purpose

Create and configure a live Google Looker Studio dashboard connected to the brand's marketing data sources. Auto-selects appropriate metrics, dimensions, and chart types based on the business model (SaaS, eCommerce, B2B, agency). Provides always-current visibility into marketing performance without manual data pulls. Eliminates the need for recurring report generation by giving stakeholders a self-service, real-time view of the metrics that matter most to their business model, with drill-down capability and date range controls built in.

Execution gate (MANDATORY — cannot be skipped)

By default this skill produces a dashboard **specification** for review. It must NOT create, publish, share, or connect a live data source to any external dashboard without passing this gate first:

1. Present the full spec — data sources, connected accounts, sharing scope, refresh cadence — as an **Execution Summary**. 2. The user must type `yes` (or an equivalent explicit approval) before any external dashboard is created or shared. ANY other input — ambiguous, implied, partial, or absent approval — cancels; the spec is saved but nothing is created externally. 3. Never proceed on ambiguous input. Never auto-retry a failed creation. 4. Record the approval with `python "${CLAUDE_PLUGIN_ROOT}/scripts/approval-manager.py" --brand {slug} --action create-approval --data '{"risk_level":"medium","summary":"..."}'` **before** creating, then `python "${CLAUDE_PLUGIN_ROOT}/scripts/approval-manager.py" --brand {slug} --action mark-executed --id {approval_id}` after it verifies.

Input Required

The user must provide (or will be prompted for):

• **Business model**: `saas` (recurring revenue focus — MRR, churn, activation, expansion), `ecommerce` (transaction focus — revenue, AOV, conversion rate, product performance), `b2b-lead-gen` (pipeline focus — MQLs, SQLs, pipeline value, CPL), or `agency` (multi-client focus — client health scores, utilization, cross-client performance). Determines the default metric set, layout template, and visualization priorities
• **Data sources to connect**: Which platforms to pull into the dashboard — Google Analytics (traffic, behavior, conversions), Google Ads (paid search performance, spend), Meta Ads (paid social performance, spend), CRM (pipeline, deal data, customer lifecycle), email platform (campaign performance, list health). Multiple sources can be combined into unified views with cross-platform calculated fields
• **Primary KPIs to feature**: The 3-5 headline metrics to display prominently at the top of the dashboard — e.g., MRR and churn rate for SaaS, revenue and ROAS for eCommerce, SQLs and pipeline value for B2B. These appear as scorecard widgets with trend indicators and target comparisons
• **Dashboard audience**: `executives` (high-level scorecards with trend arrows, minimal drill-down, focused on business outcomes), `marketing-team` (full operational detail with channel breakdowns, campaign-level data, and diagnostic dimensions), or `client` (branded presentation view with performance against stated objectives, competitive context, and clean visual design)
• **Refresh frequency**: How often the data should update — `real-time` (streaming where supported), `daily` (standard for most use cases), `weekly` (for executive dashboards with less granular needs). Determines data source caching configuration and extract schedule

Process

1. **Load brand context**: Read `~/.claude-marketing/brands/_active-brand.json` for the active slug, then load `~/.claude-marketing/brands/{slug}/profile.json`. Extract business model, key metrics, industry vertical, brand colors for dashboard theming, and connected platform credentials. Check for guidelines at `~/.claude-marketing/brands/{slug}/guidelines/_manifest.json`. If no brand exists, ask: "Set up a brand first (/digital-marketing-pro:brand-setup)?" — or proceed with defaults. 2. **Design dashboard layout based on business model template**: Select the appropriate metric hierarchy and page structure. For SaaS: page 1 overview (MRR scorecard, churn rate, CAC, LTV, CAC:LTV ratio), page 2 acquisition funnel (traffic to trial to activation to paid, by channel), page 3 retention (cohort retention curves, expansion revenue, net revenue retention). For eCommerce: page 1 overview (revenue, AOV, conversion rate, ROAS), page 2 product performance (top products, category breakdown, inventory velocity), page 3 channel mix (attributed revenue by channel, campaign-level ROAS). For B2B: page 1 overview (MQLs, SQLs, pipeline value, win rate), page 2 funnel (lead to MQL to SQL to opportunity to closed, conversion rates per stage), page 3 channel efficiency (CPL, cost per SQL, cost per opportunity by channel). For Agency: page 1 portfolio overview (client health scores, total managed spend, utilization), page 2 per-client drill-down (selectable client filter with full KPI set), page 3 cross-client benchmarks. 3. **Map data sources to dashboard widgets**: For each widget in the layout, ident