3 Commits

Author	SHA1	Message	Date
Igor Loskutov	8373874cbd	feat: add LLM_ENABLE_THINKING env var for thinking-mode LLMs ... Some LLMs (e.g. GLM-4.5-Air) default to thinking mode which returns content in reasoning_content instead of content field, breaking structured output parsing. This setting passes enable_thinking through extra_body to control the behavior per deployment. Three states: None (default, don't send), True, False.	2026-02-06 20:19:53 -05:00
Sergey Mankovsky	5f7dfadabd	fix: retry on workflow timeout (#798)	2025-12-18 20:49:06 +01:00
Igor Monadical	61f0e29d4c	feat: llm retries (#739) ... * llm retries no-mistakes * self-review (no-mistakes) * self-review (no-mistakes) * bigger retry intervals by default * tests and dry * restore to main state * parse retries * json retries (no-mistakes) * json retries (no-mistakes) * json retries (no-mistakes) * json retries (no-mistakes) self-review * additional network retry test * more lindt --------- Co-authored-by: Igor Loskutov <igor.loskutoff@gmail.com>	2025-12-05 12:08:21 -05:00