diff --git a/.github/workflows/gha_workflow_llama_stack_tests.yml b/.github/workflows/gha_workflow_llama_stack_tests.yml
index 587534311..8f7a25ee4 100644
--- a/.github/workflows/gha_workflow_llama_stack_tests.yml
+++ b/.github/workflows/gha_workflow_llama_stack_tests.yml
@@ -41,6 +41,11 @@ on:
         required: true
         default: "meta-reference"
 
+      api_key:
+        description: 'Provider API key'
+        required: false
+        default: "---"
+
 env:
   # Path to model checkpoints within EFS volume
   MODEL_CHECKPOINT_DIR: "/data/llama/Llama3.2-3B-Instruct"
@@ -52,7 +57,13 @@ env:
   MODEL_IDS: "${{ inputs.model_ids || 'Llama3.2-3B-Instruct' }}"
 
   # ID used for each test's provider config
-  PROVIDER_ID: "${{ inputs.provider_id || 'meta-reference' }}"
+  #PROVIDER_ID: "${{ inputs.provider_id || 'meta-reference' }}"
+
+  # Defined dynamically when each test is run below
+  #PROVIDER_CONFIG: ""
+
+  # (Unused) API key that can be manually defined for workflow dispatch
+  API_KEY: "${{ inputs.api_key || '' }}"
 
   # Defines which directories in TESTS_PATH to exclude from the test loop
   EXCLUDED_DIRS: "__pycache__"
@@ -67,7 +78,7 @@ jobs:
       pull-requests: write
     defaults:
       run:
-        shell: bash
+        shell: bash # default shell to run all steps for a given job.
     runs-on: ${{ inputs.runner != '' && inputs.runner || 'llama-stack-gha-runner-gpu' }}
     if: always()
     steps:
@@ -134,6 +145,14 @@ jobs:
       ############################
       #### UPDATE SYSTEM PATH ####
       ############################
+      - name: "[DEBUG] Update path: before"
+        id: path_update_before
+        if: ${{ inputs.debug == 'true' }}
+        run: |
+          echo "System path before update:"
+          echo "PATH=$PATH"
+          echo "GITHUB_PATH=$GITHUB_PATH"
+
       - name: "Update path: execute"
         id: path_update_exec
         run: |
@@ -142,6 +161,14 @@ jobs:
           mkdir -p ${HOME}/.local/bin
           echo "${HOME}/.local/bin" >> "$GITHUB_PATH"
 
+      - name: "[DEBUG] Update path: after"
+        id: path_update_after
+        if: ${{ inputs.debug == 'true' }}
+        run: |
+          echo "System path after update:"
+          echo "PATH=$PATH"
+          echo "GITHUB_PATH=$GITHUB_PATH"
+
       ##################################
       #### DEPENDENCY INSTALLATIONS ####
       ##################################
@@ -202,7 +229,6 @@ jobs:
         working-directory: "${{ github.workspace }}"
         run: |
           pattern=""
-          echo "PROVIDER_ID = ${PROVIDER_ID}"
           for dir in llama_stack/providers/tests/*; do
             if [ -d "$dir" ]; then
               dir_name=$(basename "$dir")
@@ -210,7 +236,7 @@ jobs:
                 for file in "$dir"/test_*.py; do
                   test_name=$(basename "$file")
                   new_file="result-${dir_name}-${test_name}.xml"
-                  if torchrun $(which pytest) -s -v ${TESTS_PATH}/${dir_name}/${test_name} -m "meta-reference and llama_3b" \
+                  if torchrun $(which pytest) -s -v ${TESTS_PATH}/${dir_name}/${test_name} -m "meta_reference and llama_3b" \
                      --junitxml="${{ github.workspace }}/${new_file}"; then
                     echo "Test passed: $test_name"
                   else
diff --git a/llama_stack/providers/inline/inference/meta_reference/inference.py b/llama_stack/providers/inline/inference/meta_reference/inference.py
index 07fd4af44..6743ca822 100644
--- a/llama_stack/providers/inline/inference/meta_reference/inference.py
+++ b/llama_stack/providers/inline/inference/meta_reference/inference.py
@@ -36,6 +36,8 @@ class MetaReferenceInferenceImpl(Inference, ModelRegistryHelper, ModelsProtocolP
     def __init__(self, config: MetaReferenceInferenceConfig) -> None:
         self.config = config
         model = resolve_model(config.model)
+        if model is None:
+            raise RuntimeError(f"Unknown model: {config.model}, Run `llama model list`")
         ModelRegistryHelper.__init__(
             self,
             [
@@ -45,8 +47,6 @@ class MetaReferenceInferenceImpl(Inference, ModelRegistryHelper, ModelsProtocolP
                 )
             ],
         )
-        if model is None:
-            raise RuntimeError(f"Unknown model: {config.model}, Run `llama model list`")
         self.model = model
         # verify that the checkpoint actually is for this model lol