Дашборд затрат и токенов сессии

productivity

tldr-stats is a Claude Code skill that runs the `tldr_stats.py` script and prints a full session dashboard directly in the chat, showing actual Claude API spend, token counts (sent, received, prompt cache), TLDR file-compression savings, and hook activity. The output includes progress bars for compression ratios, per-session sparklines, TLDR cache hit rates, and daemon uptime. Cost estimates use fixed per-model rates: Opus $15/1M, Sonnet $3/1M, Haiku $0.25/1M. It's the go-to skill when you want to see exactly how much a session cost, verify that TLDR and hooks are firing, or compare token usage before and after enabling file summarization.