v4 Schritt 1: optimale Kontextfenster automatisch ermitteln

- hw_math: max_ctx_for() (Umkehrung der ctx-Heuristik) + extract_params_b + recommend_ctx.
- cookbook.py: optimal_ctx pro Datei in analyze + evaluate.
- models.py status: params_b + optimal_ctx pro Modell (psutil-RAM).
- models.js Konfig-Modal: Empfehlung + 'Optimal uebernehmen'.
- cookbook.js Modal: 'Empfohlener Kontext ... uebernehmen'.

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

routers/models.py

@@ -16,8 +16,10 @@ from auth import auth
 from config import CMD_TEMPLATE, CONFIG_PATH, DEFAULT_TTL, LLAMA_SWAP_URL, MODELS_DIR, TOKEN
 from jobengine import JOBS, start_job
 from llamaswap import _swap_get, read_config, write_config
+from hw_math import extract_params_b, max_ctx_for
 import re
 import os
+import psutil
 
 router = APIRouter(prefix="/api", dependencies=[Depends(auth)])
 
@@ -54,6 +56,10 @@ class UpdateReq(BaseModel):
 @router.get("/status")
 def status():
     cfg = read_config()
+    try:
+        ram_gb = psutil.virtual_memory().total / (1024 ** 3)
+    except Exception:  # noqa: BLE001
+        ram_gb = 0
     configured = {}
     for name, spec in (cfg.get("models") or {}).items():
         spec = spec or {}
@@ -94,7 +100,9 @@ def status():
                 "size_bytes": size_bytes,
                 "quant": quant,
                 "caps": caps,
-                "filename": filename
+                "filename": filename,
+                "params_b": extract_params_b(filename or name),
+                "optimal_ctx": (max_ctx_for(extract_params_b(filename or name), quant or "Q4_K_M", ram_gb) if ram_gb else None),
             }
         }
     swap_ok = True