{"version":"2.1.285","anchor":"prompt-cache-tail-audit-telemetry","canonical_anchor":"prompt-cache-tail-audit-telemetry","heading":"Internal reporting on why parts of a request miss the prompt cache","tier":"internal","area":"Elsewhere","url":"https:\/\/changelogs.core-directive.com\/v\/2.1.285\/e\/prompt-cache-tail-audit-telemetry","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.285","markdown":"### Internal reporting on why parts of a request miss the prompt cache\n\nClaude Code now measures and reports internally why the end of a conversation is left out of the prompt cache\n\n**Unclear.** It is not clear where this reporting is sent or whether a switch controls it.\n\n**What**\n\nThe prompt cache lets Claude Code reuse parts of a request it has already sent, instead of sending them again in full. When building a request, Claude Code now checks the most recent messages that the cache does not cover. It reports the reasons and an estimate of the tokens involved; a token is a small piece of text. This is internal reporting.\n\n**Why**\n\nIt measures how often, and why, requests miss the cache.\n\n- Area: Elsewhere\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5"}