fix(providers): inject OpenRouter reasoning.effort for thinking models

Follow-up to #3851: that PR added `extra_body.thinking={type: disabled}`
for MiMo via OpenRouter, but OR doesn't forward provider-specific
thinking shapes to upstream — it strips unknown extra_body fields and
uses its own unified `reasoning` parameter. So MiMo via OR kept
thinking despite the injection (reproduced by @ClearPlume on #3851
with identical kwargs but provider switched from openrouter → xiaomi_mimo).

For known thinking-capable models (Kimi, MiMo) routed via the
openrouter spec, also inject `extra_body.reasoning = {effort: <effort>}`
in OR's documented enum ("none"|"minimal"|"low"|"medium"|"high"|"xhigh").
OR translates this to the upstream model's native shape.

Existing tests updated to expect both fields on the OR path. The direct
xiaomi_mimo and moonshot paths are unchanged (the new branch is gated
on spec.name == "openrouter"). Flash and non-MiMo models on OR continue
to receive no injection.

nanobot/providers/openai_compat_provider.py

@@ -615,6 +615,27 @@ class OpenAICompatProvider(LLMProvider):
                 {"thinking": {"type": "enabled" if thinking_enabled else "disabled"}}
             )
 
+        # OpenRouter uses its own unified `reasoning` field and does not
+        # forward provider-specific thinking shapes (the Kimi/MiMo
+        # extra_body.thinking above) to upstream. Reported as the follow-up
+        # to #3845/#3851: MiMo via OR kept thinking despite our injection.
+        # For known thinking-capable models routed via OR, mirror the
+        # effort signal into reasoning.effort (OR's documented enum:
+        # "none"|"minimal"|"low"|"medium"|"high"|"xhigh"), which OR
+        # translates to the upstream model's native shape.
+        if (
+            spec
+            and spec.name == "openrouter"
+            and reasoning_effort is not None
+            and (
+                _is_kimi_thinking_model(model_name)
+                or _is_mimo_thinking_model(model_name)
+            )
+        ):
+            kwargs.setdefault("extra_body", {}).update(
+                {"reasoning": {"effort": semantic_effort}}
+            )
+
         if tools:
             kwargs["tools"] = tools
             kwargs["tool_choice"] = tool_choice or "auto"