AI Signal 497
tare skill parses Claude Code logs to attribute token costs and diagnose quota drain
tare is a Claude Code skill that reads the client's local request logs to explain token consumption, letting users ask in plain English why they hit limits or where their quota went.
Claude Code's log format repeats each API response multiple times, inflating naive token counts by 86% on the author's test data. tare deduplicates entries and attributes context re-sending costs to the tools that caused them, giving users actionable explanations rather than raw numbers.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
tare installs with one npx command and requires no accounts or configuration beyond starting a new Claude Code session.
The tool deduplicates log entries and attributes context re-sending costs to their cause, correcting the 86% inflation from naive counting.
Everything runs locally with no network connections, and shareable summaries can be redacted to remove private information.
THE CLUSTER