fix(custom-provider): emit reasoning_effort at the live profile path

PR #57601's original branch added a top-level reasoning_effort emit to the
LEGACY build_kwargs path (agent/transports/chat_completions.py), but
provider=custom resolves to CustomProfile (plugins/model-providers/custom/),
so chat_completion_helpers takes the profile path and returns early — the
added branch was unreachable dead code for every custom endpoint.

Move the fix to its real site, CustomProfile.build_api_kwargs_extras(), and
follow the DeepSeek/Zai profile precedent:
  - disabled            -> extra_body.think = False (unchanged)
  - enabled + effort    -> TOP-LEVEL reasoning_effort (the OpenAI-compatible
                           format GLM-5.2/ARK expect), passed through verbatim
                           incl. max/xhigh
  - enabled + no effort -> omit, so the endpoint's server default applies
                           (avoids silently forcing 'medium' as the original
                           branch did)

Deliberately does NOT force think=True on enable — that flag is Ollama-only
and risks a 400 on GLM/vLLM endpoints that don't recognize it; thinking is
already server-default-on for these backends.

Verified end-to-end through the real profile dispatch (temp HERMES_HOME):
custom+high -> reasoning_effort=high; custom+max -> reasoning_effort=max;
custom+none -> think=False; custom+unset -> nothing; num_ctx composes.

Adds tests/plugins/model_providers/test_custom_profile.py (13 cases).
Addresses the custom-provider half of #55276.

Co-authored-by: huanshan5195 <huanshan5195@users.noreply.github.com>

plugins/model-providers/custom/__init__.py

@@ -1,9 +1,13 @@
 """Custom / Ollama (local) provider profile.
 
 Covers any endpoint registered as provider="custom", including local
-Ollama instances. Key quirks:
+Ollama instances and OpenAI-compatible reasoning endpoints (GLM-5.2 on
+Volcengine ARK, vLLM, llama.cpp). Key quirks:
   - ollama_num_ctx → extra_body.options.num_ctx (local context window)
   - reasoning_config disabled → extra_body.think = False
+  - reasoning_config enabled + effort → top-level reasoning_effort
+    (the native OpenAI-compatible format GLM/ARK expect; unset omits it
+    so the endpoint's server default applies)
 """
 
 from typing import Any
@@ -23,6 +27,7 @@ class CustomProfile(ProviderProfile):
         **ctx: Any,
     ) -> tuple[dict[str, Any], dict[str, Any]]:
         extra_body: dict[str, Any] = {}
+        top_level: dict[str, Any] = {}
 
         # Ollama context window
         if ollama_num_ctx:
@@ -30,14 +35,29 @@ class CustomProfile(ProviderProfile):
             options["num_ctx"] = ollama_num_ctx
             extra_body["options"] = options
 
-        # Disable thinking when reasoning is turned off
+        # Reasoning / thinking control for custom OpenAI-compatible endpoints
+        # (GLM-5.2 on Volcengine ARK, vLLM, Ollama, llama.cpp, …).
+        #
+        #   - disabled  → extra_body.think = False (Ollama's thinking-off flag)
+        #   - enabled + effort set → TOP-LEVEL reasoning_effort string, the
+        #     format GLM-5.2/ARK and other OpenAI-compatible reasoning APIs
+        #     expect (GLM documents "high" and "max"; "max" is its default).
+        #   - enabled + no effort  → omit both, so the endpoint applies its own
+        #     server-side default (do NOT force a level the user didn't pick).
+        #
+        # We deliberately do NOT emit ``think=True`` on enable: it is an
+        # Ollama-only flag and thinking is already server-default-on for these
+        # backends, so forcing it risks a 400 on GLM/vLLM endpoints that don't
+        # recognize it. Mirrors the DeepSeek/Zai profile precedent.
         if reasoning_config and isinstance(reasoning_config, dict):
             _effort = (reasoning_config.get("effort") or "").strip().lower()
             _enabled = reasoning_config.get("enabled", True)
             if _effort == "none" or _enabled is False:
                 extra_body["think"] = False
+            elif _effort:
+                top_level["reasoning_effort"] = _effort
 
-        return extra_body, {}
+        return extra_body, top_level
 
     def fetch_models(
         self,