{"version":"2.1.293","anchor":"pasted-content-pruning-in-the-input-compares-text-with-unico","canonical_anchor":"pasted-content-pruning-in-the-input-compares-text-with-unico","heading":"Pasted content in the prompt is matched more carefully as you edit","tier":"notice","area":"Prompt Input","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.293\/e\/pasted-content-pruning-in-the-input-compares-text-with-unico","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.293","markdown":"### Pasted content in the prompt is matched more carefully as you edit\n\nAs you edit the prompt, Claude Code now matches pasted content using Unicode-normalised text and caches each comparison\n\n**Unclear.** It is not clear what visible problem with pasted content in the prompt this fixes.\n\n**What**\n\nWhen you edit your prompt, Claude Code works out which pieces of pasted content still belong to it. That matching has changed. It now compares text after Unicode normalisation, so letters that look the same but are stored differently (such as an accented character written as one symbol or as a letter plus an accent) count as equal. It also checks the text each paste added to the prompt separately, and remembers the result for each pasted item instead of working it out again every time.\n\nBefore, a simpler comparison was used.\n\n**Why**\n\nPasted content should be kept or removed according to what is actually in the prompt, even when text that looks identical is encoded differently.\n\n- Area: Prompt Input\n- Tier: You'll notice\n- Useful: 1\/5\n- Signal: 0\/5\n- Scope: individual\n- Heads-up: no"}