feat(llm): Live-Parameter top_p + Live/Restart-Kennzeichnung — Plan-Schritt 4

- local_llm_top_p (Default 0.9) durch Config -> runtime_config (RUNTIME_SETTABLE,
  live ohne Neustart) -> local-openai-compatible Payload (top_p).
- Admin-Einstellungen: top_p als Slider (0–1); Hints "wirkt sofort (kein Neustart)"
  für temperature/top_p.
- Doku: top_p + Klarstellung Live-Parameter (temp/top_p/max_tokens) vs.
  Startup-Wert Kontextfenster (OLLAMA_CONTEXT_LENGTH / llama.cpp -c).

Verifiziert: top_p im Payload, Live-Config-Round-Trip (PUT/DELETE). 165 grün.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

app/providers/llm/local_openai_compatible.py

@@ -20,6 +20,7 @@ class LocalOpenAICompatibleLLM(LLMProvider):
         disable_reasoning: bool = True,
         max_tokens: int = 0,
         temperature: float = 0.3,
+        top_p: float = 0.9,
     ):
         self.base_url = base_url.rstrip("/")
         self.api_key = api_key
@@ -28,6 +29,7 @@ class LocalOpenAICompatibleLLM(LLMProvider):
         self.disable_reasoning = disable_reasoning
         self.max_tokens = max_tokens
         self.temperature = temperature
+        self.top_p = top_p
 
     def _build_messages(
         self, text: str, history: list[dict] | None, language: str | None = None
@@ -66,6 +68,7 @@ class LocalOpenAICompatibleLLM(LLMProvider):
             "model": self.model,
             "messages": self._build_messages(text, history, language=language),
             "temperature": self.temperature,
+            "top_p": self.top_p,
         }
         if stream:
             payload["stream"] = True