kpn load: Ollama-mallin vaihto lennossa (0.5b → 32b)

- Hub: uusi POST /api/v1/model endpoint, broadcastaa change_model
- Native node: kuuntelee change_model, kutsuu Ollaman pull + vaihtaa mallin
- Frontend: kpn load näyttää 5 mallia, numero vaihtaa Ollaman mallin
- Selain-WASM pysyy 0.5B:nä (kpn load 1)

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

network-poc/native-node/src/inference.rs

@@ -1,8 +1,9 @@
 use std::time::Instant;
+use std::cell::RefCell;
 
 pub struct LlmEngine {
     ollama_url: String,
-    model: String,
+    model: RefCell<String>,
     client: reqwest::Client,
 }
 
@@ -18,24 +19,29 @@ impl LlmEngine {
             .build()
             .map_err(|e| format!("HTTP client: {}", e))?;
 
-        Ok(LlmEngine { ollama_url, model, client })
+        Ok(LlmEngine { ollama_url, model: RefCell::new(model), client })
     }
 
-    pub fn model_name(&self) -> &str {
-        &self.model
+    pub fn model_name(&self) -> String {
+        self.model.borrow().clone()
+    }
+
+    pub fn set_model(&self, new_model: String) {
+        *self.model.borrow_mut() = new_model;
     }
 
     /// Varmistaa että malli on ladattu Ollamaan (ollama pull)
     pub async fn ensure_model(&self) -> Result<(), String> {
-        tracing::info!("Tarkistetaan malli {}...", self.model);
+        let model = self.model.borrow().clone();
+        tracing::info!("Tarkistetaan malli {}...", model);
         let resp = self.client.post(format!("{}/api/pull", self.ollama_url))
-            .json(&serde_json::json!({ "name": self.model, "stream": false }))
+            .json(&serde_json::json!({ "name": model, "stream": false }))
             .send()
             .await
             .map_err(|e| format!("Ollama pull: {}", e))?;
 
         if resp.status().is_success() {
-            tracing::info!("Malli {} valmis", self.model);
+            tracing::info!("Malli {} valmis", model);
             Ok(())
         } else {
             Err(format!("Ollama pull epäonnistui: {}", resp.status()))
@@ -44,11 +50,12 @@ impl LlmEngine {
 
     pub async fn generate(&self, prompt: &str, max_tokens: usize) -> Result<GenerateResult, String> {
         let system = "You are a coding assistant. Respond with ONLY code. No explanations, no markdown, no comments unless asked.";
+        let model = self.model.borrow().clone();
 
         let start = Instant::now();
         let resp = self.client.post(format!("{}/api/generate", self.ollama_url))
             .json(&serde_json::json!({
-                "model": self.model,
+                "model": model,
                 "prompt": prompt,
                 "system": system,
                 "stream": false,