Add Hermes Agent to local apps

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

@@ -134,7 +134,7 @@ 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");
 	});
 
@@ -161,6 +161,46 @@ curl -X POST "http://localhost:8000/v1/chat/completions" \\
 		expect(snippet[2].content).toContain("pi");
 	});
 
+	it("hermes-agent", async () => {
+		const { snippet: snippetFunc } = LOCAL_APPS["hermes-agent"];
+		const model: ModelData = {
+			id: "bartowski/Llama-3.2-3B-Instruct-GGUF",
+			tags: ["conversational"],
+			gguf: { total: 1, context_length: 4096, chat_template: "{% if tools %}" },
+			inference: "",
+		};
+		const snippet = snippetFunc(model);
+
+		expect(snippet[0].content).toContain(`llama-server -hf bartowski/Llama-3.2-3B-Instruct-GGUF:{{QUANT_TAG}}`);
+		expect(snippet[1].content).toContain("hermes config set model.provider custom");
+		expect(snippet[1].content).toContain("hermes config set model.base_url http://127.0.0.1:8080/v1");
+		expect(snippet[1].content).toContain(
+			"hermes config set model.default bartowski/Llama-3.2-3B-Instruct-GGUF:{{QUANT_TAG}}",
+		);
+		expect(snippet[2].content).toContain("hermes");
+	});
+
+	it("hermes-agent - mlx", async () => {
+		const { snippet: snippetFunc } = LOCAL_APPS["hermes-agent"];
+		const model: ModelData = {
+			id: "mlx-community/Llama-3.2-3B-Instruct-mlx",
+			tags: ["mlx", "conversational"],
+			pipeline_tag: "text-generation",
+			config: {
+				tokenizer_config: {
+					chat_template: "{% if tools %}...{% endif %}",
+				},
+			},
+			inference: "",
+		};
+		const snippet = snippetFunc(model);
+
+		expect(snippet[0].setup).toContain("uv tool install mlx-lm");
+		expect(snippet[1].content).toContain("hermes config set model.provider custom");
+		expect(snippet[1].content).toContain("hermes config set model.default mlx-community/Llama-3.2-3B-Instruct-mlx");
+		expect(snippet[2].content).toContain("hermes");
+	});
+
 	it("docker model runner", async () => {
 		const { snippet: snippetFunc } = LOCAL_APPS["docker-model-runner"];
 		const model: ModelData = {

packages/tasks/src/local-apps.ts

@@ -139,6 +139,14 @@ function isUnslothModel(model: ModelData) {
 	return model.tags.includes("unsloth") || isLlamaCppGgufModel(model);
 }
 
+function isToolCallingLocalAgentModel(model: ModelData): boolean {
+	return (
+		(isLlamaCppGgufModel(model) || isMlxModel(model)) &&
+		model.tags.includes("conversational") &&
+		!!getChatTemplate(model)?.includes("tools")
+	);
+}
+
 function getQuantTag(filepath?: string): string {
 	const defaultTag = ":{{QUANT_TAG}}";
 
@@ -463,12 +471,8 @@ const snippetMlxLm = (model: ModelData): LocalAppSnippet[] => {
 	];
 };
 
-const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
-	const modelName = model.id.split("/").pop() ?? model.id;
-	const isMLX = isMlxModel(model);
-
-	// Step 1: Server — differs by backend
-	const serverStep: LocalAppSnippet = isMLX
+const getLocalServerStep = (model: ModelData, filepath?: string): LocalAppSnippet => {
+	return isMlxModel(model)
 		? {
 				title: "Start the MLX server",
 				setup: "# Install MLX LM:\nuv tool install mlx-lm",
@@ -479,16 +483,21 @@ const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
 				setup: "# Install llama.cpp:\nbrew install llama.cpp",
 				content: `# Start a local OpenAI-compatible server:\nllama-server -hf ${model.id}${getQuantTag(filepath)}`,
 			};
+};
+
+const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
+	const isMLX = isMlxModel(model);
+	const modelId = isMLX ? model.id : `${model.id}${getQuantTag(filepath)}`;
+	const serverStep = getLocalServerStep(model, filepath);
 
-	// Step 2: Pi config — port and provider name differ
 	const modelsJson = JSON.stringify(
 		{
 			providers: {
 				[isMLX ? "mlx-lm" : "llama-cpp"]: {
 					baseUrl: "http://localhost:8080/v1",
 					api: "openai-completions",
 					apiKey: "none",
-					models: [{ id: isMLX ? model.id : modelName }],
+					models: [{ id: modelId }],
 				},
 			},
 		},
@@ -510,6 +519,33 @@ const snippetPi = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
 	];
 };
 
+const snippetHermesAgent = (model: ModelData, filepath?: string): LocalAppSnippet[] => {
+	const modelId = isMlxModel(model) ? model.id : `${model.id}${getQuantTag(filepath)}`;
+	const serverStep = getLocalServerStep(model, filepath);
+
+	return [
+		serverStep,
+		{
+			title: "Configure Hermes",
+			setup: [
+				"# Install Hermes:",
+				"curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash",
+				"hermes setup",
+			].join("\n"),
+			content: [
+				"# Point Hermes at the local server:",
+				"hermes config set model.provider custom",
+				"hermes config set model.base_url http://127.0.0.1:8080/v1",
+				`hermes config set model.default ${modelId}`,
+			].join("\n"),
+		},
+		{
+			title: "Run Hermes",
+			content: "hermes",
+		},
+	];
+};
+
 const snippetDockerModelRunner = (model: ModelData, filepath?: string): string => {
 	// Only add quant tag for GGUF models, not safetensors
 	const quantTag = isLlamaCppGgufModel(model) ? getQuantTag(filepath) : "";
@@ -755,12 +791,16 @@ export const LOCAL_APPS = {
 		prettyLabel: "Pi",
 		docsUrl: "https://github.com/badlogic/pi-mono",
 		mainTask: "text-generation",
-		displayOnModelPage: (model) =>
-			(isLlamaCppGgufModel(model) || isMlxModel(model)) &&
-			model.tags.includes("conversational") &&
-			!!getChatTemplate(model)?.includes("tools"),
+		displayOnModelPage: isToolCallingLocalAgentModel,
 		snippet: snippetPi,
 	},
+	"hermes-agent": {
+		prettyLabel: "Hermes Agent",
+		docsUrl: "https://hermes-agent.nousresearch.com/",
+		mainTask: "text-generation",
+		displayOnModelPage: isToolCallingLocalAgentModel,
+		snippet: snippetHermesAgent,
+	},
 } satisfies Record<string, LocalApp>;
 
 export type LocalAppKey = keyof typeof LOCAL_APPS;