{"version":"2.1.321","anchor":"prompt-cache-normalization-gets-an-auditedrows-setting","canonical_anchor":"prompt-cache-normalization-gets-an-auditedrows-setting","heading":"The message-preparation cache gains an auditedRows setting","tier":"internal","area":"Context","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.321\/e\/prompt-cache-normalization-gets-an-auditedrows-setting","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.321","markdown":"### The message-preparation cache gains an auditedRows setting\n\nThe cache Claude Code uses to reuse earlier message preparation now has an extra auditedRows setting beside its existing verify check\n\n**What**\n\nClaude Code can reuse earlier work when it tidies the conversation into a standard form before each request to the model. That reuse now has one more setting, `auditedRows`. Claude Code takes its value from configuration and falls back to a built-in default when none is given.\n\nThe setting sits next to the existing safety check, which you can switch on with the environment variable `CLAUDE_CODE_NORMALIZE_PREFIX_CACHE_VERIFY` (an environment variable is a named setting passed to a program from your shell).\n\n**Why**\n\nThe checking side of this cache is still being worked on. Nothing changes in how you use Claude Code.\n\n- Area: Context\n- Tier: Under the hood\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}