{"version":"2.1.318","anchor":"stop-sequence-turns-after-an-api-error-are-rewritten-to-tool","canonical_anchor":"stop-sequence-turns-after-an-api-error-are-rewritten-to-tool","heading":"Claude keeps working after some API errors instead of ending the turn","tier":"notice","area":"API Errors","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.318\/e\/stop-sequence-turns-after-an-api-error-are-rewritten-to-tool","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.318","markdown":"### Claude keeps working after some API errors instead of ending the turn\n\nAfter an API error that stopped on a stop sequence, Claude Code now treats the turn as waiting for a tool, so Claude keeps going\n\n**What**\n\nA turn is one round of Claude working on a request. When a turn ends with an API error message, Claude Code now checks whether Claude should keep going. It changes the turn's ending from a stop sequence to a tool call, so the turn continues, when all of these are true:\n\n- the turn needs a follow-up\n\n- no tool asked for the turn to end\n\n- nothing has blocked continuing\n\n- no tool was put off until later\n\n**Why**\n\nA turn interrupted by an API error can carry on instead of stopping early.\n\n- Area: API Errors\n- Tier: You'll notice\n- Useful: 3\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}