Summarize a sprint's worth of Claude Code activity — sessions grouped by project (cwd), per-model cost breakdown, token efficiency (cache hit rate, compaction baselines), subagent effectiveness from workflow API, velocity metrics (turn_count, turn_duration_ms), and tool diversity across the sprint.
Summarize a sprint's worth of Claude Code activity — sessions grouped by project (cwd), per-model cost breakdown, token efficiency (cache hit rate, compaction baselines), subagent effectiveness from workflow API, velocity metrics (turn_count, turn_duration_ms), and tool diversity across the sprint.