ci(e2e): revert PR 5600 key routing

.github/workflows/e2e-script.yaml

@@ -62,11 +62,6 @@ on:
         required: false
         type: boolean
         default: false
-      nvidia_api_key_alias:
-        description: Emit NVIDIA_API_KEY compatibility alias for documented legacy shell consumers.
-        required: false
-        type: boolean
-        default: false
       brave_api_key:
         description: Pass the BRAVE_API_KEY secret to the script.
         required: false
@@ -85,8 +80,6 @@ on:
     secrets:
       NVIDIA_INFERENCE_API_KEY:
         required: false
-      NVIDIA_API_KEY:
-        required: false
       BRAVE_API_KEY:
         required: false
       DOCKERHUB_USERNAME:
@@ -150,7 +143,6 @@ jobs:
         with:
           ref: ${{ github.ref }}
           sparse-checkout: |
-            .github/actions/export-e2e-hosted-inference
             .github/actions/run-e2e-script
           path: workflow-actions
           persist-credentials: false
@@ -222,11 +214,26 @@ jobs:
 
       - name: Export hosted CI inference environment
         if: ${{ inputs.nvidia_api_key }}
-        uses: ./workflow-actions/.github/actions/export-e2e-hosted-inference
-        with:
-          nvidia-inference-api-key: ${{ secrets.NVIDIA_INFERENCE_API_KEY }}
-          nvidia-api-key: ${{ secrets.NVIDIA_API_KEY }}
-          export-nvidia-api-key: ${{ inputs.nvidia_api_key_alias }}
+        env:
+          NVIDIA_INFERENCE_API_KEY: ${{ secrets.NVIDIA_INFERENCE_API_KEY }}
+        shell: bash
+        run: |
+          set -euo pipefail
+
+          if [ -z "${NVIDIA_INFERENCE_API_KEY:-}" ]; then
+            echo "::error::NVIDIA_INFERENCE_API_KEY secret is required for hosted CI inference; it is withheld for workflow_dispatch target_ref runs." >&2
+            exit 1
+          fi
+
+          {
+            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/nvidia/nemotron-3-super-v3\n'
+            printf 'NEMOCLAW_COMPAT_MODEL=nvidia/nvidia/nemotron-3-super-v3\n'
+            printf 'NEMOCLAW_PREFERRED_API=openai-completions\n'
+            printf 'COMPATIBLE_API_KEY=%s\n' "${NVIDIA_INFERENCE_API_KEY}"
+          } >> "$GITHUB_ENV"
 
       - name: Run E2E script
         uses: ./workflow-actions/.github/actions/run-e2e-script
@@ -241,6 +248,7 @@ jobs:
         env:
           BRAVE_API_KEY: ${{ inputs.brave_api_key && secrets.BRAVE_API_KEY || '' }}
           GITHUB_TOKEN: ${{ inputs.github_token && github.token || '' }}
+          NVIDIA_INFERENCE_API_KEY: ${{ inputs.nvidia_api_key && secrets.NVIDIA_INFERENCE_API_KEY || '' }}
           TELEGRAM_BOT_TOKEN_REAL: ${{ inputs.messaging_live_secrets && secrets.TELEGRAM_BOT_TOKEN_REAL || '' }}
           TELEGRAM_CHAT_ID_E2E: ${{ inputs.messaging_live_secrets && secrets.TELEGRAM_CHAT_ID_E2E || '' }}
           DISCORD_BOT_TOKEN_REAL: ${{ inputs.messaging_live_secrets && secrets.DISCORD_BOT_TOKEN_REAL || '' }}