{"version":"2.1.314","anchor":"reactive-compaction-can-tell-the-summarizer-about-left-out-t","canonical_anchor":"reactive-compaction-can-tell-the-summarizer-about-left-out-t","heading":"Conversation summaries can note tool results that were left out","tier":"notice","area":"Compaction","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314\/e\/reactive-compaction-can-tell-the-summarizer-about-left-out-t","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314","markdown":"### Conversation summaries can note tool results that were left out\n\nWhen Claude Code summarizes a long conversation to free up room, it can now tell the summarizer which tool results were left out\n\n**Unclear.** It is not clear what makes Claude Code include the note about left-out tool results in a summary.\n\n**What**\n\nCompaction is how Claude Code shrinks a long conversation by replacing earlier parts with a summary, so there is room to keep going. Reactive compaction is the kind that happens as a response during the session rather than on request.\n\nThis kind of compaction can now tell the model writing the summary about tool results that were left out. A tool result is the output of an action Claude Code took, such as reading a file. In some cases it also adds extra wording to the summarizer's instructions.\n\n**Why**\n\nThe way conversations are summarized has changed. The summary can now reflect that some tool output was not included.\n\n- Area: Compaction\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}