Prompts

Act as my revenue analyst. Prepare the monthly revenue review for [month and year].

1. Call get_revenue_kpis and report MRR, ARR, ARPA, net revenue retention, CAC and customer counts, each with its change from the previous month.
2. Call get_mrr_movements for the same period and explain what moved MRR: new, expansion, contraction, churn and reactivation.
3. Call get_missed_revenue and report any revenue missed to failed or missing payments.
4. Call get_revenue_goals, then get_revenue_goal_progress for each active goal, and say whether we are ahead or behind.

Format: a one-paragraph summary first, then a table of KPIs, then three bullets on risks and one on what to do next.
Rules: state the currency once. Use only numbers the tools returned. If a figure is missing, say so rather than estimating.

Explain the change in MRR over [time window, e.g. the last 3 months].

1. Call get_mrr_movements and show new, expansion, contraction, churn and reactivation MRR per month.
2. For the two largest movements in the most recent month, call list_mrr_movement_customers and name the top 5 customers behind each, with plan and amount.
3. Say in two sentences what drove the net change.

Rules: do not guess reasons for churn. Only describe what the data shows, such as plan, amount and date.

Find revenue we are losing to missed payments.

1. Call get_missed_revenue for the last 6 months and show expected versus missed revenue per month.
2. For the most recent month, call list_missed_payment_customers and list the customers ordered by MRR at risk, with status and days overdue.
3. For the 3 largest, find their billing customer ids with list_revenue_account_mappings, then call get_account_payment_calendar for this year and tell me whether this is a one-off or a pattern.

Format: a short summary with the total MRR at risk, then the table.
Rules: do not contact or change anything. This is a read-only review.

Review our plan mix.

1. Call get_plan_mrr_summary and show each plan's customers, MRR, ARR and share of customers and MRR.
2. Call get_plan_mrr_trend for the three plans with the most MRR and show how each moved over the last 6 months.
3. Call get_acv_summary and compare average contract value across plans.

Then tell me: which plan carries our revenue, which is growing fastest, and which has many customers but little MRR.
Rules: separate what the data shows from your own interpretation, and label each.

Check whether we are on track for our revenue goals.

1. Call get_revenue_goals and list each active goal with its target and dates.
2. For each, call get_revenue_goal_progress and report expected versus actual for the current month and the variance.
3. Call get_revenue_kpis to confirm the current value of each metric.

Then rank the goals from most behind to most ahead, and say in one sentence what would close the biggest gap.
Rules: if there are no active goals, say so and suggest which metric to set one for.

Review our marketing channels for the last 30 days.

1. Call get_channel_performance and show visitors, sessions, identified users and accounts per channel.
2. Call get_multi_touch_attribution and compare the revenue each channel gets under first-touch, last-touch and linear attribution.
3. Call get_campaign_performance for the top 2 channels and name the best and worst campaigns.
4. Call get_campaign_revenue_ranking for the best channel and list the campaigns ranked by revenue.

Format: a table of channels, then three recommendations on where to spend more or less.
Rules: use only returned numbers, and flag any channel where the attribution models disagree strongly.

Check the quality of our recent signups.

1. Call get_new_signup_counts and report new users and accounts for the last day, 7 days and 30 days with their change.
2. Call get_lead_quality for the last month and give the mix of corporate, personal-email and abusive or fake leads.
3. Call list_leads_by_quality for corporate leads from the last 7 days and list the 10 newest.
4. Call get_acquisition_timing and tell me how many touches and days a visitor takes to sign up.

End with whether quality is improving, and which source to look at first.
Rules: do not email or contact any lead.

Review activation and retention for [month and year].

1. Call get_activation_summary and get_activation_trend and report milestone completion and the trend.
2. Call get_feature_conversion_drivers (step modules), pick the activation module with the weakest milestone, then call it again (step milestone) and compare converted and dropped accounts.
3. Call get_retention_data and describe how cohorts return over time.
4. Call get_churn_expansion_signals and report the high, medium and low stickiness split.

Format: findings first, then the one milestone I should work on and why.
Rules: separate what the data shows from your own interpretation.

Build a funnel scorecard for the last 12 months.

1. Call get_monthly_metrics_table (it takes no inputs) and show it as a table, month by month.
2. Call get_mau_mtu_count for the most recent complete month (as YYYY-MM) and add active and tracked users.
3. Point out the month-to-month step with the biggest drop and the one with the biggest gain.

Rules: do not smooth or round away a drop. Say which figures are missing, if any.

Prepare our weekly at-risk account review.

1. Call list_at_risk_accounts and take the 5 accounts with the most at stake.
2. For each, call get_account_overview and get_account_engagement_trend, and say whether the risk is engagement, monetization or both.
3. For each, call get_account_feature_usage and say which features they use and how widely.
4. For accounts at risk on monetization, find their billing customer ids and check get_account_payment_calendar for missed payments.

Format: one short paragraph per account, ending with a suggested next step.
Rules: this is read-only. Do not message anyone.

Give me a brief on [account name].

1. Find the account with list_crm_accounts. If there is more than one match, ask me which one.
2. Call get_account_overview, then get_account_acquisition to see how they found us.
3. Call get_account_users and get_account_activity for the last 30 days: who is active and what they do.
4. Find the account's billing customer ids with list_revenue_account_mappings and call get_account_revenue for MRR, plan and payment status.

Format: six bullets I can read in under a minute, then anything I should follow up on.
Rules: if the account has no linked billing customer, say so and skip step 4.

Check whether our ThriveStack data can be trusted.

1. Call get_setup_status and get_marketing_setup_status and report what is connected and what is missing.
2. Call get_revenue_sync_status and report the latest sync per source, with status and time. Flag anything failed or stale.
3. Call list_revenue_customer_mappings with filter UNMAPPED and report how many billing customers are not linked to an account, and the biggest few.
4. Call list_revenue_account_mappings with filter UNMAPPED and report how many product accounts have no billing customer.

End with a verdict: safe to report, safe with caveats (list them), or not yet.
Rules: do not fix or change anything.