fix(hosts): gate host aliases on docker/vm drivers (#4516)

src/lib/actions/sandbox/host-aliases.ts

@@ -5,10 +5,18 @@ import { isIP } from "node:net";
 
 import { dockerExecFileSync } from "../../adapters/docker";
 import { CLI_NAME } from "../../cli/branding";
+import * as registry from "../../state/registry";
 
 const K3S_CONTAINER = "openshell-cluster-nemoclaw";
 const HOST_ALIAS_KUBECTL_TIMEOUT_MS = 10_000;
 
+// Drivers that run a per-sandbox direct container (openshell-<sandbox>...)
+// instead of the legacy k3s gateway. They have no openshell-cluster-nemoclaw
+// container and no Kubernetes `Sandbox` custom resource, so the kubectl-based
+// host-alias backend below cannot target them. See src/lib/sandbox/privileged-exec.ts
+// for the direct-container resolution these drivers use elsewhere.
+const DIRECT_CONTAINER_DRIVERS = new Set(["docker", "vm"]);
+
 type HostAlias = {
   ip: string;
   hostnames: string[];
@@ -57,6 +65,33 @@ function hostAliasesFail(lines: string | readonly string[], exitCode = 1): never
   throw new HostAliasesCommandError(lines, exitCode);
 }
 
+function normalizeDriver(driver: unknown): string | null {
+  return typeof driver === "string" && driver.trim()
+    ? driver.trim().toLowerCase()
+    : null;
+}
+
+// Host aliases are persisted on the legacy Kubernetes gateway `Sandbox`
+// custom resource and applied by `docker exec openshell-cluster-nemoclaw
+// kubectl ...`. The docker and vm drivers run per-sandbox direct containers
+// with no gateway cluster container and no `Sandbox` CR, so this k3s
+// control-plane path cannot work for them. Fail fast with an actionable
+// message instead of targeting a container that does not exist (#4516) — and
+// without pretending a one-time /etc/hosts edit inside the direct container
+// would survive a sandbox restart or rebuild.
+function assertLegacyGatewayHostAliasSupport(sandboxName: string): void {
+  const driver = normalizeDriver(registry.getSandbox(sandboxName)?.openshellDriver);
+  if (driver && DIRECT_CONTAINER_DRIVERS.has(driver)) {
+    hostAliasesFail([
+      `  Host aliases are not supported on the '${driver}' driver sandbox '${sandboxName}'.`,
+      "  This command edits aliases on the legacy Kubernetes gateway sandbox resource,",
+      `  which the ${driver} driver does not run (there is no openshell-cluster-nemoclaw container).`,
+      "  OpenShell does not yet expose a persistent host-alias API for this driver, and a",
+      "  one-time /etc/hosts edit would not survive a sandbox restart or rebuild.",
+    ]);
+  }
+}
+
 function validateHostAliasHostname(hostname: string): boolean {
   if (!hostname || hostname.length > 253) return false;
   return hostname.split(".").every((label) => {
@@ -186,6 +221,7 @@ function patchHostAliasesWithRetry(
 }
 
 export function listSandboxHostAliases(sandboxName: string): void {
+  assertLegacyGatewayHostAliasSupport(sandboxName);
   const aliases = getHostAliases(getSandboxResource(sandboxName));
   if (aliases.length === 0) {
     console.log(`  No host aliases configured for '${sandboxName}'.`);
@@ -207,6 +243,7 @@ export function addSandboxHostAlias(
   sandboxName: string,
   options: AddSandboxHostAliasOptions = {},
 ): void {
+  assertLegacyGatewayHostAliasSupport(sandboxName);
   const dryRun = Boolean(options.dryRun);
   const { hostname: rawHostname, ip } = options;
   if (!rawHostname || !ip) {
@@ -249,6 +286,7 @@ export function removeSandboxHostAlias(
   sandboxName: string,
   options: RemoveSandboxHostAliasOptions = {},
 ): void {
+  assertLegacyGatewayHostAliasSupport(sandboxName);
   const dryRun = Boolean(options.dryRun);
   const { hostname: rawHostname } = options;
   if (!rawHostname) {

test/cli.test.ts

@@ -140,6 +140,7 @@ type SandboxEntry = {
   gpuEnabled: boolean;
   policies: string[];
   agent?: string;
+  openshellDriver?: string | null;
   agentVersion?: string | null;
 };
 
@@ -2582,6 +2583,47 @@ describe("CLI dispatch", () => {
     });
   });
 
+  for (const driver of ["docker", "vm"] as const) {
+    it(`gates host alias commands on the ${driver} driver without targeting the legacy gateway container`, testTimeoutOptions(30_000), () => {
+      const home = fs.mkdtempSync(
+        path.join(os.tmpdir(), `nemoclaw-cli-hosts-${driver}-`),
+      );
+      const localBin = path.join(home, "bin");
+      const dockerLog = path.join(home, "docker.log");
+      fs.mkdirSync(localBin, { recursive: true });
+      // Record any docker invocation so we can prove the gate fires before
+      // the legacy `docker exec openshell-cluster-nemoclaw kubectl` path.
+      writeHostAliasDockerStub(localBin, dockerLog, [
+        { ip: "10.0.0.5", hostnames: ["old.local"] },
+      ]);
+      writeSandboxRegistry(home, "alpha", { openshellDriver: driver });
+
+      const env = { HOME: home, PATH: `${localBin}:${process.env.PATH || ""}` };
+      const list = runWithEnv("alpha hosts-list", env);
+      const add = runWithEnv("alpha hosts-add searxng.local 192.168.1.105", env);
+      const remove = runWithEnv("alpha hosts-remove searxng.local", env);
+
+      for (const result of [list, add, remove]) {
+        expect(result.code).toBe(1);
+        expect(result.out).toContain(
+          `Host aliases are not supported on the '${driver}' driver sandbox 'alpha'.`,
+        );
+      }
+
+      // Even the dry-run preview must not reach the legacy resource read.
+      const dryRun = runWithEnv(
+        "alpha hosts-add searxng.local 192.168.1.105 --dry-run",
+        env,
+      );
+      expect(dryRun.code).toBe(1);
+      expect(dryRun.out).not.toContain("/spec/podTemplate/spec/hostAliases");
+
+      // The gate runs before any docker exec, so the legacy gateway container
+      // is never targeted.
+      expect(fs.existsSync(dockerLog)).toBe(false);
+    });
+  }
+
   it("supports oclif-native sandbox command forms", () => {
     const home = fs.mkdtempSync(path.join(os.tmpdir(), "nemoclaw-cli-native-sandbox-"));
     writeSandboxRegistry(home);