{"version":"2.1.300","anchor":"up-front-unattended-serving-consent-question-for-remote-cont","canonical_anchor":"up-front-unattended-serving-consent-question-for-remote-cont","heading":"Remote Control can ask up front whether to allow unattended tool calls","tier":"notice","area":"Remote Control","scope":"individual","heads_up":true,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.300\/e\/up-front-unattended-serving-consent-question-for-remote-cont","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.300","markdown":"### Remote Control can ask up front whether to allow unattended tool calls\n\nBefore Remote Control starts, Claude Code can now ask once whether to allow tool calls while you are away, unless `--allow-unattended-tool-calls` is passed\n\n**Unclear.** It is not clear what starts this question or whether it is shown when Remote Control starts in this release.\n\n**What**\n\nRemote Control lets you drive a Claude Code session from another device. Claude Code can now show a consent question before Remote Control starts, asking whether to allow unattended tool calls, meaning actions the model takes while nobody is watching.\n\n- The answer is remembered, so the question is asked once.\n\n- The question is skipped when you pass `--allow-unattended-tool-calls`, or when there is no interactive terminal.\n\n- The question times out if nobody answers.\n\n- If saving the answer fails, Claude Code says so.\n\nThe existing consent dialog gained the timeout and save-failure handling to support this.\n\n**Why**\n\nThis adds an interactive prompt to the Remote Control startup path, which matters if you start it from a script.\n\n- Area: Remote Control\n- Tier: You'll notice\n- Useful: 3\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: yes"}