{"version":"2.1.314","anchor":"sandbox-note-replaces-sandboxed-failure-note","canonical_anchor":"sandbox-note-replaces-sandboxed-failure-note","heading":"Sandbox note now shown on every sandboxed command and again after compaction","tier":"notice","area":"Sandbox","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314\/e\/sandbox-note-replaces-sandboxed-failure-note","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314","markdown":"### Sandbox note now shown on every sandboxed command and again after compaction\n\nThe model is now told about sandbox limits on sandboxed commands, not only on failures, and is told again after a conversation is compacted\n\n**What**\n\nA sandbox is a restricted space that limits what a command can reach, such as your files or the network. Claude Code tells the model about these limits through a note attached to the result of a tool call (an action the model takes, like running a command).\n\n- A new \"sandboxed\" note replaces the old `sandboxed_failure` note. It now appears on sandboxed commands in general, not only when they fail.\n\n- The note says whether the command was limited for both files and network (`files_and_network`) or for network only (`network_only`).\n\n- The first time in a conversation, the note is shown in a slightly different wording.\n\n- When a conversation is compacted (shortened by summarising it), the flag that records the note was already shown, `toldOfSandbox`, is reset, so the note can appear again.\n\n- The record kept for a compaction turn now also includes `responseBeforeCompaction`.\n\n**Why**\n\nThe model is told about sandbox limits more consistently, including after compaction, so it is less likely to be surprised when a command cannot reach a file or the network.\n\n- Area: Sandbox\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}