feat: add Fireworks AI provider for chat agent

Routes chat-completions through a managed OpenAI-compatible inference
endpoint as an alternative to local Ollama, useful when the agent needs
fast multi-hop reasoning that local hardware can't sustain.

- backend: rename `call_ollama_chat_messages` → `call_chat_messages`,
  dispatch by provider; add `call_fireworks` branch (Bearer auth,
  `response_format: json_object` mapped from internal `format="json"`)
  and `list_fireworks_models` Tauri command
- settings: extend `AiProvider` enum + `AiSettings.fireworks_api_key`
  (serde-default for legacy config compat); Fireworks base URL hardcoded
- UI: provider selector in both popover and AppSettingsSheet (only
  ollama+fireworks shown; legacy openai/anthropic kept for serde-compat
  but normalized to ollama in UI); password input + dynamic model list
  for Fireworks; switching provider clears stale model selection
- 4 unit tests: serde round-trip, legacy settings deserialization,
  Fireworks chat-completions parsing, models-list parsing

src-tauri/src/models/ai.rs

@@ -7,14 +7,19 @@ pub enum AiProvider {
     Ollama,
     OpenAi,
     Anthropic,
+    Fireworks,
 }
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct AiSettings {
     pub provider: AiProvider,
     pub ollama_url: String,
+    #[serde(default)]
     pub openai_api_key: Option<String>,
+    #[serde(default)]
     pub anthropic_api_key: Option<String>,
+    #[serde(default)]
+    pub fireworks_api_key: Option<String>,
     pub model: String,
 }
 
@@ -25,11 +30,14 @@ impl Default for AiSettings {
             ollama_url: "http://localhost:11434".to_string(),
             openai_api_key: None,
             anthropic_api_key: None,
+            fireworks_api_key: None,
             model: String::new(),
         }
     }
 }
 
+/// Generic chat message used by all chat providers (Ollama, Fireworks, OpenAI-compatible).
+/// `{role, content}` shape is identical across these APIs.
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct OllamaChatMessage {
     pub role: String,
@@ -55,7 +63,48 @@ pub struct OllamaTagsResponse {
     pub models: Vec<OllamaModel>,
 }
 
+/// Generic chat-model descriptor exposed to the UI dropdown.
+/// Reused as the return shape for both Ollama and Fireworks model listings.
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct OllamaModel {
     pub name: String,
 }
+
+// ---------------------------------------------------------------------------
+// Fireworks (OpenAI-compatible chat-completions)
+// ---------------------------------------------------------------------------
+
+#[derive(Debug, Clone, Serialize)]
+pub struct FireworksChatRequest {
+    pub model: String,
+    pub messages: Vec<OllamaChatMessage>,
+    pub temperature: f32,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub response_format: Option<FireworksResponseFormat>,
+}
+
+#[derive(Debug, Clone, Serialize)]
+pub struct FireworksResponseFormat {
+    #[serde(rename = "type")]
+    pub kind: String,
+}
+
+#[derive(Debug, Deserialize)]
+pub struct FireworksChatResponse {
+    pub choices: Vec<FireworksChoice>,
+}
+
+#[derive(Debug, Deserialize)]
+pub struct FireworksChoice {
+    pub message: OllamaChatMessage,
+}
+
+#[derive(Debug, Deserialize)]
+pub struct FireworksModelsResponse {
+    pub data: Vec<FireworksModelEntry>,
+}
+
+#[derive(Debug, Deserialize)]
+pub struct FireworksModelEntry {
+    pub id: String,
+}