3 commits

Author	SHA1	Message	Date
Mario Zechner	d327b9c768	fix(ai): handle same-provider different-model handoff in OpenAI Responses API ... When switching between OpenAI models (e.g., gpt-5-mini to gpt-5.2-codex), function_call IDs with fc_ prefix trigger pairing validation errors because OpenAI tracks which fc_xxx IDs were paired with rs_xxx reasoning items. The fix omits the id field for function_calls from different models, which avoids the pairing validation while keeping call_id for matching with function_call_output. Fixes #886	2026-01-22 00:58:49 +01:00
Mario Zechner	de58391085	test(ai): add failing test for orphaned function_call without reasoning item ... Reproduces issue #886 where function_call is sent without its required paired reasoning item, causing Azure/OpenAI 400 error.	2026-01-21 23:59:29 +01:00
Mario Zechner	2d27a2c728	fix(ai): skip errored/aborted assistant messages in transform-messages ... Fixes OpenAI Responses 400 error 'reasoning without following item' by skipping errored/aborted assistant messages entirely rather than filtering at the provider level. This covers openai-responses, openai-codex-responses, and future providers. Removes strictResponsesPairing compat option (no longer needed). Closes #838	2026-01-19 15:55:29 +01:00