Score a session against the rolling population average and report its percentile on\ncost, tokens, tool count, and complexity using Agent Monitor data.
\nThe user provides: $ARGUMENTS
\nThis may be:
\n| Endpoint | \nReturns | \n
|---|---|
GET /api/sessions?limit=N | \nPopulation of sessions with cost, model, started_at, metadata (turn_count, total_turn_duration_ms) — builds the rolling baseline | \n
GET /api/pricing/cost/{sessionId} | \n{ total_cost, breakdown:[{ input_tokens, output_tokens, cache_read_tokens, cache_write_tokens, cost }] } — the target session's cost and tokens | \n
GET /api/workflows/{sessionId} | \ncomplexity (score), stats (tool/event counts), toolFlow (distinct tools used) — the target session's tool count and complexity | \n
GET /api/analytics | \navg_events_per_session, tool_usage, daily_sessions — corroborates population-level averages | \n
Fetch the population with GET /api/sessions?limit=200 (the rolling set). For each\nsession gather cost (GET /api/pricing/cost/{id} or the list cost field), total\ntokens (sum of the 4 token types from the pricing breakdown), tool count and\ncomplexity (GET /api/workflows/{id}). Compute mean, median, and standard\ndeviation for each metric across the population.
For the requested session, pull the same four metrics:
\ntotal_cost from GET /api/pricing/cost/{id}.input + output + cache_read + cache_write summed from the breakdown.GET /api/workflows/{id} stats/toolFlow.complexity.score from GET /api/workflows/{id}.For each metric report the target's percentile within the population (share of\nsessions at or below it) and its z-score (value − mean) / stddev. Label each:\nbelow average / typical / above average / outlier (|z| > 2).
State whether the session was normal overall. If it is an outlier, name which\nmetric drove it (e.g., complexity p96, cost p91 → an unusually heavy session).
\n