Update to ViewsMax

{
  "description": "Summarize how the user's ViewsMax offers and tracked links are performing — video views, clicks, calls booked, email sign-ups, sales revenue, and daily click and revenue trends for a date range. Use for questions like \"how did my offers do last month\", \"which offer made the most money\", or \"is traffic going up\".",
  "included_files": [],
  "name": "report-offer-performance",
  "skill_md_contents": "---\nname: report-offer-performance\ndescription: Summarize how the user's ViewsMax offers and tracked links are performing — video views, clicks, calls booked, email sign-ups, sales revenue, and daily click and revenue trends for a date range. Use for questions like \"how did my offers do last month\", \"which offer made the most money\", or \"is traffic going up\".\n---\n\n# Report offer performance from ViewsMax\n\n## 1. Pin down the window\n\nUse the dates the user gives. If they don't give any, use the last 28 days and\nsay so. Send `from`/`to` as ISO-8601 dates.\n\n## 2. Pull the numbers\n\n1. Call `get_offer_stats` for the window to get totals: video views, clicks,\n   calls booked, email sign-ups, and sales revenue.\n2. Call `get_stats_timeseries` for the same window to get daily clicks and\n   attributed revenue. Add `event_id` when the user asks about one offer.\n3. Call `list_offers` when the user wants a per-offer or per-link breakdown.\n   Its results include each offer's links and stats.\n\n## 3. Present it\n\nLead with a short answer to the user's question, then show:\n\n- **Totals** for the window.\n- **Top offers or links**, ranked by the metric the user cares about. If they\n  didn't say, rank by revenue, then by clicks.\n- **Trend:** compare the first and second half of the window, or week over\n  week. Name the best and worst days.\n- **Next step:** one or two concrete suggestions that follow from the numbers,\n  such as moving a high-converting link to a busier placement.\n\n## Accuracy rules\n\n- Report only metrics the tools returned. If you calculate a rate, such as\n  revenue per click, show the formula and label it as calculated.\n- Zero-filled days in the time series mean no recorded activity. They don't\n  mean missing data.\n- If every metric is empty, say tracking may not be installed or links may not\n  be live yet, and point the user to the track-offer skill. Don't guess why.\n"
}