Command
Use list_projects({}) to confirm my supplied project host is visible. Use my chosen days value of 7, 30, or 90 for every time-windowed call. Read get_overview({host, days}) and get_gap_analysis({host, days}). Compare returned competitor observations with my saved baseline export and explicit surge threshold. Investigate candidates through list_prompts({host, limit: 100}) and get_prompt_runs({host, prompt_id, days}). Require comparable coverage before labeling a surge; otherwise mark it unverified. Draft an alert containing the rival, observed difference, supporting samples, and recommended investigation. Return a no-alert result when criteria fail. Daily scheduling and Slack delivery need an independent connection. Return evidence links, coverage limits, and explicit unknowns; never infer unsupported scores.What you need
- A visible host, supported window, saved baseline, competitor names, and surge thresholds.
- An independent scheduler and Slack connection if alerts should run or be delivered automatically.
Require evidence before calling a surge
Check whether the same competitor and measurement coverage appear in both the current observations and the supplied baseline. Review relevant answer samples before escalating, and distinguish a genuine comparable difference from a change caused by missing observations or altered coverage.
Give reviewers enough context to act
An alert draft names the competitor, shows the observed difference, and links the supporting evidence. Include the threshold used and a suggested next investigation, allowing a reviewer to approve escalation or decide that the available information does not justify an alert.