fix(onboard): preserve hosted Inference Hub model IDs

.github/workflows/e2e-script.yaml

@@ -229,8 +229,8 @@ jobs:
             printf 'NEMOCLAW_E2E_USE_HOSTED_INFERENCE=1\n'
             printf 'NEMOCLAW_PROVIDER=custom\n'
             printf 'NEMOCLAW_ENDPOINT_URL=https://inference-api.nvidia.com/v1\n'
-            printf 'NEMOCLAW_MODEL=nvidia/nemotron-3-ultra\n'
-            printf 'NEMOCLAW_COMPAT_MODEL=nvidia/nemotron-3-ultra\n'
+            printf 'NEMOCLAW_MODEL=nvidia/nvidia/nemotron-3-ultra\n'
+            printf 'NEMOCLAW_COMPAT_MODEL=nvidia/nvidia/nemotron-3-ultra\n'
             printf 'NEMOCLAW_PREFERRED_API=openai-completions\n'
             printf 'COMPATIBLE_API_KEY=%s\n' "${NVIDIA_INFERENCE_API_KEY}"
           } >> "$GITHUB_ENV"

.github/workflows/e2e-vitest-scenarios.yaml

@@ -1415,8 +1415,8 @@ jobs:
           NVIDIA_INFERENCE_API_KEY: ${{ secrets.NVIDIA_INFERENCE_API_KEY }}
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
         run: |
           set -euo pipefail
@@ -2135,8 +2135,8 @@ jobs:
       NEMOCLAW_NON_INTERACTIVE: "1"
       NEMOCLAW_PROVIDER: custom
       NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-      NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-      NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+      NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+      NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
       NEMOCLAW_PREFERRED_API: openai-completions
       NEMOCLAW_SANDBOX_NAME: e2e-rebuild-hermes
       OPENSHELL_GATEWAY: nemoclaw
@@ -2218,8 +2218,8 @@ jobs:
       NEMOCLAW_NON_INTERACTIVE: "1"
       NEMOCLAW_PROVIDER: custom
       NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-      NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-      NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+      NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+      NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
       NEMOCLAW_PREFERRED_API: openai-completions
       NEMOCLAW_SANDBOX_NAME: e2e-rebuild-hermes-base
       OPENSHELL_GATEWAY: nemoclaw
@@ -2970,8 +2970,8 @@ jobs:
       NEMOCLAW_SANDBOX_NAME: "e2e-resume"
       NEMOCLAW_PROVIDER: "custom"
       NEMOCLAW_ENDPOINT_URL: "https://inference-api.nvidia.com/v1"
-      NEMOCLAW_MODEL: "nvidia/nemotron-3-ultra"
-      NEMOCLAW_COMPAT_MODEL: "nvidia/nemotron-3-ultra"
+      NEMOCLAW_MODEL: "nvidia/nvidia/nemotron-3-ultra"
+      NEMOCLAW_COMPAT_MODEL: "nvidia/nvidia/nemotron-3-ultra"
     steps:
       - uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
         with:

.github/workflows/nightly-e2e.yaml

@@ -468,8 +468,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -544,8 +544,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_ISSUE_4434_LIVE: "1"
@@ -1003,8 +1003,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1301,8 +1301,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1608,8 +1608,8 @@ jobs:
           NVIDIA_INFERENCE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           E2E_ARTIFACT_DIR: ${{ github.workspace }}/e2e-artifacts/vitest/credential-migration
@@ -1834,8 +1834,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1847,8 +1847,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1886,8 +1886,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1899,8 +1899,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1938,8 +1938,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1951,8 +1951,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -1991,8 +1991,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2004,8 +2004,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2044,8 +2044,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2058,8 +2058,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2100,8 +2100,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2114,8 +2114,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"
@@ -2194,8 +2194,8 @@ jobs:
           NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1"
           NEMOCLAW_PROVIDER: custom
           NEMOCLAW_ENDPOINT_URL: https://inference-api.nvidia.com/v1
-          NEMOCLAW_MODEL: nvidia/nemotron-3-ultra
-          NEMOCLAW_COMPAT_MODEL: nvidia/nemotron-3-ultra
+          NEMOCLAW_MODEL: nvidia/nvidia/nemotron-3-ultra
+          NEMOCLAW_COMPAT_MODEL: nvidia/nvidia/nemotron-3-ultra
           NEMOCLAW_PREFERRED_API: openai-completions
           COMPATIBLE_API_KEY: ${{ (github.event_name != 'workflow_dispatch' || inputs.target_ref == '') && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           NEMOCLAW_NON_INTERACTIVE: "1"

src/lib/inference/onboard-probes.test.ts

@@ -293,8 +293,8 @@ describe("OpenAI-compatible inference probes", () => {
   });
 
   it("bounds the hosted compatible inference probe for the served Nemotron model", () => {
-    expect(getChatCompletionsProbePayload("nvidia/nemotron-3-ultra")).toEqual({
-      model: "nvidia/nemotron-3-ultra",
+    expect(getChatCompletionsProbePayload("nvidia/nvidia/nemotron-3-ultra")).toEqual({
+      model: "nvidia/nvidia/nemotron-3-ultra",
       messages: [{ role: "user", content: "Reply with exactly: OK" }],
       max_tokens: 8,
     });

src/lib/onboard/providers.ts

@@ -27,7 +27,10 @@ const HOSTED_INFERENCE_SOURCE_ENV = "NVIDIA_INFERENCE_API_KEY";
 const HOSTED_INFERENCE_PROVIDER_KEY_ENV = "NEMOCLAW_PROVIDER_KEY";
 const HOSTED_INFERENCE_CREDENTIAL_ENV = "COMPATIBLE_API_KEY";
 const HOSTED_INFERENCE_ENDPOINT_URL = "https://inference-api.nvidia.com/v1";
-const HOSTED_INFERENCE_MODEL = "nvidia/nemotron-3-ultra";
+// Private CI-compatible Inference Hub endpoint model IDs use the
+// provider/namespace/model convention. This endpoint is staged as a custom
+// OpenAI-compatible provider, not as the public build.nvidia.com provider.
+const HOSTED_INFERENCE_MODEL = "nvidia/nvidia/nemotron-3-ultra";
 const NON_INTERACTIVE_PROVIDER_ALIASES = {
   cloud: "build",
   nim: "nim-local",

test/e2e-scenario/fixtures/hosted-inference.ts

@@ -6,7 +6,7 @@ const HOSTED_INFERENCE_CREDENTIAL_ENV = "COMPATIBLE_API_KEY";
 const HOSTED_INFERENCE_PROVIDER = "custom";
 const HOSTED_INFERENCE_PROVIDER_NAME = "compatible-endpoint";
 const DEFAULT_HOSTED_INFERENCE_BASE_URL = "https://inference-api.nvidia.com/v1";
-const DEFAULT_HOSTED_INFERENCE_MODEL = "nvidia/nemotron-3-ultra";
+const DEFAULT_HOSTED_INFERENCE_MODEL = "nvidia/nvidia/nemotron-3-ultra";
 
 export interface HostedInferenceSecrets {
   required(name: string): string;

test/e2e-scenario/live/agent-turn-latency-helpers.ts

@@ -23,7 +23,7 @@ export const HERMES_SANDBOX =
 validateSandboxName(OPENCLAW_SANDBOX);
 validateSandboxName(HERMES_SANDBOX);
 const DEFAULT_NVIDIA_MODEL = "nvidia/nemotron-3-super-120b-a12b";
-const DEFAULT_COMPAT_MODEL = "nvidia/nemotron-3-ultra";
+const DEFAULT_COMPAT_MODEL = "nvidia/nvidia/nemotron-3-ultra";
 const USE_COMPATIBLE_HOSTED = process.env.NEMOCLAW_E2E_USE_HOSTED_INFERENCE === "1";
 export const MODEL =
   process.env.NEMOCLAW_TURN_LATENCY_MODEL ??

test/e2e-scenario/live/hermes-discord.test.ts

@@ -47,11 +47,11 @@ function commandEnv(apiKey?: string, extra: NodeJS.ProcessEnv = {}): NodeJS.Proc
       NEMOCLAW_PROVIDER: process.env.NEMOCLAW_PROVIDER ?? "custom",
       NEMOCLAW_ENDPOINT_URL:
         process.env.NEMOCLAW_ENDPOINT_URL ?? "https://inference-api.nvidia.com/v1",
-      NEMOCLAW_MODEL: process.env.NEMOCLAW_MODEL ?? "nvidia/nemotron-3-ultra",
+      NEMOCLAW_MODEL: process.env.NEMOCLAW_MODEL ?? "nvidia/nvidia/nemotron-3-ultra",
       NEMOCLAW_COMPAT_MODEL:
         process.env.NEMOCLAW_COMPAT_MODEL ??
         process.env.NEMOCLAW_MODEL ??
-        "nvidia/nemotron-3-ultra",
+        "nvidia/nvidia/nemotron-3-ultra",
       NEMOCLAW_PREFERRED_API: process.env.NEMOCLAW_PREFERRED_API ?? "openai-completions",
       DISCORD_BOT_TOKEN: DISCORD_TOKEN,
       DISCORD_SERVER_IDS,

test/e2e-scenario/live/hermes-slack-e2e-helpers.ts

@@ -40,11 +40,11 @@ function hermesSlackEnv(apiKey?: string): NodeJS.ProcessEnv {
     apiKey,
     extra: {
       ...(apiKey ? { COMPATIBLE_API_KEY: apiKey } : {}),
-      NEMOCLAW_COMPAT_MODEL: process.env.NEMOCLAW_COMPAT_MODEL ?? "nvidia/nemotron-3-ultra",
+      NEMOCLAW_COMPAT_MODEL: process.env.NEMOCLAW_COMPAT_MODEL ?? "nvidia/nvidia/nemotron-3-ultra",
       NEMOCLAW_E2E_USE_HOSTED_INFERENCE: "1",
       NEMOCLAW_ENDPOINT_URL:
         process.env.NEMOCLAW_ENDPOINT_URL ?? "https://inference-api.nvidia.com/v1",
-      NEMOCLAW_MODEL: process.env.NEMOCLAW_MODEL ?? "nvidia/nemotron-3-ultra",
+      NEMOCLAW_MODEL: process.env.NEMOCLAW_MODEL ?? "nvidia/nvidia/nemotron-3-ultra",
       NEMOCLAW_POLICY_TIER: process.env.NEMOCLAW_POLICY_TIER ?? "open",
       NEMOCLAW_PREFERRED_API: process.env.NEMOCLAW_PREFERRED_API ?? "openai-completions",
       NEMOCLAW_PROVIDER: process.env.NEMOCLAW_PROVIDER ?? "custom",

test/e2e-scenario/live/rebuild-hermes.test.ts

@@ -54,7 +54,9 @@ const BACKUP_ROOT = path.join(os.homedir(), ".nemoclaw", "rebuild-backups");
 const HOSTED_ENDPOINT_URL =
   process.env.NEMOCLAW_ENDPOINT_URL ?? "https://inference-api.nvidia.com/v1";
 const HOSTED_MODEL =
-  process.env.NEMOCLAW_MODEL ?? process.env.NEMOCLAW_COMPAT_MODEL ?? "nvidia/nemotron-3-ultra";
+  process.env.NEMOCLAW_MODEL ??
+  process.env.NEMOCLAW_COMPAT_MODEL ??
+  "nvidia/nvidia/nemotron-3-ultra";
 const OLD_BASE_TAG = `nemoclaw-hermes-old-base:${SANDBOX_NAME.toLowerCase().replace(/[^a-z0-9_.-]+/g, "-")}`;
 const CURRENT_BASE_TAG = "ghcr.io/nvidia/nemoclaw/hermes-sandbox-base:latest";
 

test/e2e-scenario/live/upgrade-stale-sandbox-helpers.ts

@@ -122,7 +122,7 @@ export function writeStaleRegistryEntry(): void {
     (typeof session.model === "string" && session.model) ||
     process.env.NEMOCLAW_MODEL ||
     process.env.NEMOCLAW_COMPAT_MODEL ||
-    "nvidia/nemotron-3-ultra";
+    "nvidia/nvidia/nemotron-3-ultra";
   const registry = readJsonFile<{
     sandboxes?: Record<string, Record<string, unknown>>;
     defaultSandbox?: string;