Files
hermes-agent/plugins
xxxigm 31f0336da7 fix(custom): omit Ollama-only think=false on strict OpenAI-compat endpoints
reasoning_effort: none was injecting extra_body.think=false for every
custom provider. Mistral (and other extra=forbid hosts) reject that
field with HTTP 422. Keep think=false on Ollama URLs only; still send
top-level reasoning_effort=none so /v1 thinking-off keeps working.
2026-08-29 21:58:00 +05:30
..