fix(coding-agent): name the llama.cpp server and cause when it cannot be reached (#11779)

When `/login llama.cpp` could not connect to the server URL, pi showed
only `Error: Failed to save API key for llama.cpp: fetch failed`. That
happens on a refused connection, a wrong port, or `https://` against a
plain-HTTP server. Node's `fetch` keeps the real reason in
`error.cause`, and pi dropped it, so the user had nothing to go on.

`LlamaClient` now sends its requests, including the `/models/sse` watch,
through a small `fetchServer` helper. When `fetch` fails at the network
level, the helper throws `Could not reach llama.cpp server at <url>:
<cause>` and keeps the original error as `cause`. Aborts and timeouts
are rethrown unchanged. HTTP error responses are handled as before.

The same message also improves `/llama` and the model catalog refresh,
because they use the same client.

**Testing**
- Added a test in `test/llama-extension.test.ts` that calls
`LlamaClient.list()` against a closed local port and expects the server
URL and `connect ECONNREFUSED ...` in the error. It fails without the
change (`got 'fetch failed'`) and passes with it. The whole file passes
(16 tests).
- `npm run check` passes.
- Manual check, `/login llama.cpp` with `http://127.0.0.1:8099` (nothing
listening):
  - Before: `Error: Failed to save API key for llama.cpp: fetch failed`
- After: `Error: Failed to save API key for llama.cpp: Could not reach
llama.cpp server at http://127.0.0.1:8099: connect ECONNREFUSED
127.0.0.1:8099`

Fixes #11745

Requested by @badlogic

Co-authored-by: triagent <triagent@localhost>

packages/coding-agent/src/extensions/llama/client.ts

@@ -158,6 +158,13 @@ export function llamaInferenceUrl(serverUrl: string): string {
 	return `${normalizeLlamaServerUrl(serverUrl)}/v1`;
 }
 
+function connectionErrorDetail(error: unknown): string {
+	const cause = error instanceof Error ? error.cause : undefined;
+	if (cause instanceof Error && cause.message) return cause.message;
+	if (cause && typeof cause === "object" && "code" in cause && typeof cause.code === "string") return cause.code;
+	return error instanceof Error ? error.message : String(error);
+}
+
 export class LlamaClient {
 	readonly serverUrl: string;
 	private readonly apiKey: string | undefined;
@@ -167,13 +174,25 @@ export class LlamaClient {
 		this.apiKey = apiKey;
 	}
 
+	/** Fetches from the server, naming the server and the network-level cause when the connection fails. */
+	private async fetchServer(path: string, init: RequestInit): Promise<Response> {
+		try {
+			return await fetch(`${this.serverUrl}${path}`, init);
+		} catch (error) {
+			if (init.signal?.aborted) throw error;
+			throw new Error(`Could not reach llama.cpp server at ${this.serverUrl}: ${connectionErrorDetail(error)}`, {
+				cause: error,
+			});
+		}
+	}
+
 	private async request(path: string, init: RequestInit = {}): Promise<unknown> {
 		const headers = new Headers(init.headers);
 		if (init.body !== undefined) headers.set("Content-Type", "application/json");
 		if (this.apiKey) headers.set("Authorization", `Bearer ${this.apiKey}`);
 		const timeout = AbortSignal.timeout(15_000);
 		const signal = init.signal ? AbortSignal.any([init.signal, timeout]) : timeout;
-		const response = await fetch(`${this.serverUrl}${path}`, { ...init, headers, signal });
+		const response = await this.fetchServer(path, { ...init, headers, signal });
 		let payload: unknown;
 		try {
 			payload = await response.json();
@@ -229,7 +248,7 @@ export class LlamaClient {
 	async watch(onEvent: (event: LlamaModelEvent) => void, signal?: AbortSignal): Promise<void> {
 		const headers = new Headers();
 		if (this.apiKey) headers.set("Authorization", `Bearer ${this.apiKey}`);
-		const response = await fetch(`${this.serverUrl}/models/sse`, { headers, signal });
+		const response = await this.fetchServer("/models/sse", { headers, signal });
 		if (!response.ok || !response.body) throw new Error(`llama.cpp SSE returned HTTP ${response.status}`);
 		const reader = response.body.getReader();
 		const decoder = new TextDecoder();

packages/coding-agent/test/llama-extension.test.ts

@@ -59,6 +59,19 @@ describe("llama.cpp extension", () => {
 		expect(() => normalizeLlamaServerUrl("file:///tmp/llama")).toThrow("http or https");
 	});
 
+	// #11745: a refused connection showed only "fetch failed"
+	it("names the server and the cause when it cannot be reached", async () => {
+		const server = createServer();
+		server.listen(0, "127.0.0.1");
+		await once(server, "listening");
+		const port = (server.address() as AddressInfo).port;
+		await new Promise<void>((resolve) => server.close(() => resolve()));
+
+		await expect(new LlamaClient(`http://127.0.0.1:${port}`).list()).rejects.toThrow(
+			`Could not reach llama.cpp server at http://127.0.0.1:${port}: connect ECONNREFUSED 127.0.0.1:${port}`,
+		);
+	});
+
 	it("exposes loaded and sleeping models with router metadata", () => {
 		const controller = createLlamaProvider();
 		controller.setCatalog(