fix(embed): fast-fail + auto-restart Ollama on solved-search hang

Embeddings timeout was 120s, causing multi-minute hangs when Ollama was
down during F4/F6/F7 solved-search. Now: connect_timeout=5s across all
HTTP clients (REST + Ollama), read_timeout reduced to 30s for embeddings,
and OllamaEmbeddings auto-restarts the server on ConnectionError before
degrading gracefully.

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

harness/tht/config.py

@@ -52,6 +52,7 @@ class RestConfig(BaseModel):
     base_url: str
     api_key: str
     timeout: int = 30
+    connect_timeout: int = 5
     ssl_ca: str | None = None  # path al certificato CA (per server con CA interna)
 
 
@@ -105,7 +106,8 @@ class EmbeddingsConfig(BaseModel):
     model: str = "nomic-embed-text-v2-moe"
     dim: int = 768
     batch_size: int = 32
-    timeout: int = 120
+    timeout: int = 30
+    connect_timeout: int = 5
     bin: str = "ollama"
     start_cmd: list[str] | None = None