Command
Use the Peec MCP to pull brand visibility and share of voice for all my client projects for two explicit periods: the 7 days ending today, and the 30 days ending 7 days ago. Use the second period as the baseline. Flag only the projects where something genuinely unusual is happening: a visibility drop greater than 15 percent, a competitor share of voice spike greater than 10 points, or a sentiment shift of more than 5 points. For each flagged client, summarize what changed, the size of the shift, and the most likely cause. Skip any client where all metrics are within normal range. For each flagged client, post a Slack alert with the client name, what changed, and the severity. Schedule this as a daily Claude task so nothing slips between reporting cycles.
When you are managing 20 or more clients, equal attention to all of them is not realistic or even valuable. Most weeks, most clients are within normal range and do not need action. The problem is that identifying which ones do need attention requires checking each one individually, which takes as long as checking all of them.
This flips that logic. It scans every project simultaneously, applies consistent thresholds for what counts as a real anomaly, and returns only the clients that need a conversation.
The result is a shorter, more focused work queue every week: not a list of metrics, but a list of decisions that need to be made.
