Skill v1.0.0

name: dba-user-activity-analysis description: Use when asked to analyze Teradata user activity over the past week — ranks users by resource consumption and drills into the heaviest consumers' recent SQL. metadata: version: "1.0.0"

You are analyzing Teradata user activity for the past 7 days (or whatever window the user asks for), using a connected teradata-mcp-server.

Workflow

1. System-wide resource usage. Call dba_resusageSummary(no_days=7). A single call returns rows already broken down by LogDate, hourOfDay, dayOfWeek, workloadType, workloadComplexity, UserName, and AppID — you don't need to call it more than once or once per dimension.
2. Rank users. From the step 1 results, rank users by CPU time (primary), IO operations (secondary), and memory/byte usage (tertiary). Compute each user's share of total system resources.
3. Drill into the top consumers. For each of the top 5 users by CPU time, call dba_userSqlList(user_name=..., no_days=7) to get their recent SQL activity. user_name is required — this tool cannot be called for "all users" at once, so call it once per user of interest.
4. Report. Produce a dashboard-style write-up:

• System-wide resource overview with an hour-of-day heatmap for the period, color-coded by CPU usage.
• The user ranking table from step 2, with resource shares as percentages.
• Each top user's peak activity hours and workload mix.
• For the top 5: their recent SQL activity, which tables they hit most, and how complex/long their queries ran.
• Sortable tables and color-coded metrics throughout; include specific query and table examples where useful.

Gotchas

• dba_userSqlList requires a specific user_name — you cannot call it with no arguments or an empty name to get "everyone's" SQL history. Use dba_resusageSummary's per-user breakdown for the system-wide view, and only call dba_userSqlList per user when you need that user's individual SQL history.