Rank people by activity
Rank people by the points their activity earns, all time or over the last days, and give each person their level, like a leaderboard
Give each thing people did ({person, kind, at}), the points each kind is worth (weights), the counts to keep (metrics: {key, label, lookback_days}, null for all time), the levels ({level, name, min_points}) and which count ranks them (rank_by). Returns each person with their rank, points, every count and level, best first.
What it asks for
- activity (Needed, list): What each person did and when
- weights (Needed, details): Points per kind of activity, like {"post": 5, "reply": 1}
- metrics (Needed, list):
- levels (Needed, list):
- rank_by (Optional, words): Optional: the metric key to rank by (default the first metric)
- now (Optional, words): Optional: the time the windows count back from (ISO; default now)
Use it from your AI
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.
Say to your AI: “Use Trillion to rank people by the points their activity earns, all time or over the last days, and give each person their level, like a leaderboard”
What your AI calls: run_tool {"tool_id":"tool:rank-people-by-activity"}