{"version":"2.1.312","anchor":"ccr-worker-keeps-api-error-request-ids-in-uploaded-transcrip","canonical_anchor":"ccr-worker-keeps-api-error-request-ids-in-uploaded-transcrip","heading":"Remote sessions keep request IDs on API errors in uploaded transcripts","tier":"internal","area":"Sessions","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.312\/e\/ccr-worker-keeps-api-error-request-ids-in-uploaded-transcrip","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.312","markdown":"### Remote sessions keep request IDs on API errors in uploaded transcripts\n\nIn remote sessions, API error messages in the uploaded transcript can now keep their request ID, which makes failures easier to trace\n\n**Unclear.** It is not clear what decides whether the request ID is kept, so it may not apply to every remote session.\n\n**What**\n\nIn a remote session, Claude Code uploads the transcript, which is the record of the conversation. Before upload it strips some wrapper details from each message. API error messages can now keep their request ID through that step. A request ID is the identifier of a single call to Anthropic's servers.\n\n**Why**\n\nWith the request ID kept, an API error in a remote session can be matched to the exact call that failed, which helps when you report or look into a failure.\n\n- Area: Sessions\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}