server: add gpt-oss chat template for proper prompt formatting

The gpt-oss model requires a specific prompt format with <|start|>,
<|message|>, <|end|>, <|channel|> tokens. Without this, the model
produces degenerate output. Auto-detected via config.model_type.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

crates/xserv-server/src/main.rs

@@ -11,6 +11,7 @@ use xserv_model::ModelConfig;
 
 pub struct AppState {
     pub model_name: String,
+    pub model_type: String,
     pub engine_sender: Mutex<mpsc::Sender<GenerateRequest>>,
     pub engine_tokenizer: Mutex<xserv_tokenizer::Tokenizer>,
     pub max_seq_len: usize,
@@ -99,8 +100,10 @@ async fn main() {
         }
     });
 
+    let model_type = model_config.model_type.clone().unwrap_or_default();
     let state = Arc::new(AppState {
         model_name,
+        model_type,
         engine_sender: Mutex::new(tx),
         engine_tokenizer: Mutex::new(tokenizer),
         max_seq_len,