{"version":"2.1.311","anchor":"effort-slider-now-checks-permission-mode-and-prompt-cache-wa","canonical_anchor":"effort-slider-now-checks-permission-mode-and-prompt-cache-wa","heading":"Effort-change confirmation now considers permission mode and prompt cache","tier":"notice","area":"Effort","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.311\/e\/effort-slider-now-checks-permission-mode-and-prompt-cache-wa","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.311","markdown":"### Effort-change confirmation now considers permission mode and prompt cache\n\nBefore confirming a change of effort level, Claude Code now also weighs the permission mode and whether the prompt cache may still be warm\n\n**What**\n\nWhen you change the effort level (how much thinking Claude puts into its answers), Claude Code can ask you to confirm first. That confirmation step now takes more into account than before:\n\n- the permission mode, which is the setting for how freely Claude Code may act without asking you\n\n- the conversation so far\n\n- whether the prompt cache may still be warm. The prompt cache is a store of recent conversation that lets the next request be handled more cheaply and quickly; \"warm\" means it is still holding your recent work.\n\nBefore, it looked only at the new effort level, the current one, the model, whether you had acknowledged the token cost, and the conversation.\n\n**Why**\n\nThe warning that switching effort can lose the benefit of the prompt cache may now appear in different situations than before.\n\n- Area: Effort\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}