fix: Restore all files lost during destructive rebase

A previous `git pull --rebase origin main` dropped 177 local commits,
losing 3400+ files across admin-v2, backend, studio-v2, website,
klausur-service, and many other services. The partial restore attempt
(660295e2) only recovered some files.

This commit restores all missing files from pre-rebase ref 98933f5e
while preserving post-rebase additions (night-scheduler, night-mode UI,
NightModeWidget dashboard integration).

Restored features include:
- AI Module Sidebar (FAB), OCR Labeling, OCR Compare
- GPU Dashboard, RAG Pipeline, Magic Help
- Klausur-Korrektur (8 files), Abitur-Archiv (5+ files)
- Companion, Zeugnisse-Crawler, Screen Flow
- Full backend, studio-v2, website, klausur-service
- All compliance SDKs, agent-core, voice-service
- CI/CD configs, documentation, scripts

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/llm_gateway/routes/chat.py

@@ -0,0 +1,112 @@
+"""
+Chat Completions Route - OpenAI-kompatible API.
+"""
+
+import logging
+import json
+from typing import AsyncIterator
+from fastapi import APIRouter, HTTPException, Depends
+from fastapi.responses import StreamingResponse
+
+from ..models.chat import (
+    ChatCompletionRequest,
+    ChatCompletionResponse,
+    ChatMessage,
+    ModelListResponse,
+)
+from ..services.inference import get_inference_service, InferenceService
+from ..services.playbook_service import get_playbook_service, PlaybookService
+from ..middleware.auth import verify_api_key
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter(tags=["LLM"])
+
+
+def get_services():
+    """Dependency für Services."""
+    return get_inference_service(), get_playbook_service()
+
+
+@router.post("/chat/completions", response_model=ChatCompletionResponse)
+async def chat_completions(
+    request: ChatCompletionRequest,
+    _: str = Depends(verify_api_key),
+):
+    """
+    OpenAI-kompatible Chat Completions.
+
+    Unterstützt:
+    - Streaming (stream=true)
+    - Playbook-basierte System Prompts (metadata.playbook_id)
+    - Multiple Models (breakpilot-teacher-8b, claude-3-5-sonnet, etc.)
+    """
+    inference_service, playbook_service = get_services()
+
+    # Playbook System Prompt injizieren
+    if request.metadata and request.metadata.playbook_id:
+        playbook = playbook_service.get_playbook(request.metadata.playbook_id)
+        if playbook:
+            # System Prompt an den Anfang der Messages einfügen
+            system_msg = ChatMessage(role="system", content=playbook.system_prompt)
+            # Prüfen ob bereits ein System Prompt existiert
+            has_system = any(m.role == "system" for m in request.messages)
+            if not has_system:
+                request.messages.insert(0, system_msg)
+            else:
+                # Playbook Prompt vor bestehenden System Prompt setzen
+                for i, msg in enumerate(request.messages):
+                    if msg.role == "system":
+                        msg.content = f"{playbook.system_prompt}\n\n{msg.content}"
+                        break
+
+    try:
+        if request.stream:
+            return StreamingResponse(
+                stream_response(request, inference_service),
+                media_type="text/event-stream",
+                headers={
+                    "Cache-Control": "no-cache",
+                    "Connection": "keep-alive",
+                    "X-Accel-Buffering": "no",
+                },
+            )
+        else:
+            response = await inference_service.complete(request)
+            return response
+
+    except ValueError as e:
+        logger.error(f"Chat completion error: {e}")
+        raise HTTPException(status_code=400, detail=str(e))
+    except Exception as e:
+        logger.exception(f"Chat completion failed: {e}")
+        raise HTTPException(status_code=500, detail="Internal server error")
+
+
+async def stream_response(
+    request: ChatCompletionRequest,
+    inference_service: InferenceService,
+) -> AsyncIterator[str]:
+    """Generator für SSE Streaming."""
+    try:
+        async for chunk in inference_service.stream(request):
+            data = chunk.model_dump_json()
+            yield f"data: {data}\n\n"
+        yield "data: [DONE]\n\n"
+    except Exception as e:
+        logger.exception(f"Streaming error: {e}")
+        error_data = json.dumps({"error": str(e)})
+        yield f"data: {error_data}\n\n"
+
+
+@router.get("/models", response_model=ModelListResponse)
+async def list_models(
+    _: str = Depends(verify_api_key),
+):
+    """
+    Liste verfügbarer Modelle.
+
+    Gibt alle konfigurierten Modelle zurück, die aktuell verfügbar sind.
+    """
+    inference_service = get_inference_service()
+    return await inference_service.list_models()