Pi: use full repo id + quant tag in models.json

Matches the form Hermes uses (org/repo:QUANT_TAG) — clearer than the
bare basename and identical at runtime since llama-server ignores the
model field.

Author: gary149

packages/tasks/src/local-apps.spec.ts

@@ -138,7 +138,7 @@ curl -X POST "http://localhost:8000/v1/chat/completions" \\
 
 		expect(snippet[0].content).toContain(`llama-server -hf bartowski/Llama-3.2-3B-Instruct-GGUF:{{QUANT_TAG}}`);
 		expect(snippet[1].setup).toContain("npm install -g @mariozechner/pi-coding-agent");
-		expect(snippet[1].content).toContain(`"id": "Llama-3.2-3B-Instruct-GGUF"`);
+		expect(snippet[1].content).toContain(`"id": "bartowski/Llama-3.2-3B-Instruct-GGUF:{{QUANT_TAG}}"`);
 		expect(snippet[2].content).toContain("pi");
 	});
 

packages/tasks/src/local-apps.ts

@@ -486,8 +486,8 @@ const getLocalServerStep = (model: ModelData, filepath?: string): LocalAppSnippe
 };
 
 const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
-	const modelName = model.id.split("/").pop() ?? model.id;
 	const isMLX = isMlxModel(model);
+	const modelId = isMLX ? model.id : `${model.id}${getQuantTag(filepath)}`;
 	const serverStep = getLocalServerStep(model, filepath);
 
 	const modelsJson = JSON.stringify(
@@ -497,7 +497,7 @@ const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
 					baseUrl: "http://localhost:8080/v1",
 					api: "openai-completions",
 					apiKey: "none",
-					models: [{ id: isMLX ? model.id : modelName }],
+					models: [{ id: modelId }],
 				},
 			},
 		},