feat(ci): use replay mode, setup ollama if specific label exists on PR (#2955)

This PR makes setting up Ollama optional for CI. By default, we use
`replay` mode for inference requests and use the stored results from the
`tests/integration/recordings/` directory.

Every so often, users will update tests which will need us to re-record.
To do this, we check for the existence of a label `re-record-tests` on
the PR. If detected,
- ollama is spun up
- inference mode is set to record
- after the tests are done, if any new changes are detected, they are
pushed back to the PR

## Test Plan

This is GitHub CI. Gotta test it live.

.github/workflows/integration-tests.yml

@@ -7,6 +7,7 @@ on:
     branches: [ main ]
   pull_request:
     branches: [ main ]
+    types: [opened, synchronize, reopened, labeled, unlabeled]
     paths:
       - 'llama_stack/**'
       - 'tests/**'
@@ -39,6 +40,8 @@ jobs:
     runs-on: ubuntu-latest
     outputs:
       test-type: ${{ steps.generate-matrix.outputs.test-type }}
+      rerecord-tests: ${{ steps.check-rerecord-tests.outputs.rerecord-tests }}
+
     steps:
       - name: Checkout repository
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
@@ -52,10 +55,30 @@ jobs:
             sort | jq -R -s -c 'split("\n")[:-1]')
           echo "test-type=$TEST_TYPES" >> $GITHUB_OUTPUT
 
+      - name: Check if re-record-tests label exists
+        id: check-rerecord-tests
+        run: |
+          if [[ "${{ contains(github.event.pull_request.labels.*.name, 're-record-tests') }}" == "true" ]]; then
+            echo "rerecord-tests=true" >> $GITHUB_OUTPUT
+          else
+            echo "rerecord-tests=false" >> $GITHUB_OUTPUT
+          fi
+
   test-matrix:
     needs: discover-tests
     runs-on: ubuntu-latest
 
+    permissions:
+      # Set write permissions since we might need to commit recordings
+      contents: write
+      pull-requests: write
+
+    env:
+      # Create reusable variable for the re-record tests condition
+      SHOULD_RECORD: ${{ needs.discover-tests.outputs.rerecord-tests == 'true' }}
+      # TODO: set up another var to track whether we need ollama or not
+      # not every matrix type needs ollama
+
     strategy:
       fail-fast: false
       matrix:
@@ -74,6 +97,16 @@ jobs:
             test-type: tool_runtime
 
     steps:
+      - name: Debug
+        run: |
+          echo "test-type: ${{ matrix.test-type }}"
+          echo "client-type: ${{ matrix.client-type }}"
+          echo "provider: ${{ matrix.provider }}"
+          echo "python-version: ${{ matrix.python-version }}"
+          echo "client-version: ${{ matrix.client-version }}"
+          echo "SHOULD_RECORD: ${{ env.SHOULD_RECORD }}"
+          echo "rerecord-tests: ${{ needs.discover-tests.outputs.rerecord-tests }}"
+
       - name: Checkout repository
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
 
@@ -84,7 +117,7 @@ jobs:
           client-version: ${{ matrix.client-version }}
 
       - name: Setup ollama
-        if: ${{ matrix.provider == 'ollama' }}
+        if: ${{ matrix.provider == 'ollama' && env.SHOULD_RECORD == 'true' }}
         uses: ./.github/actions/setup-ollama
 
       - name: Setup vllm
@@ -116,6 +149,14 @@ jobs:
           fi
 
           EXCLUDE_TESTS="builtin_tool or safety_with_image or code_interpreter or test_rag"
+          export LLAMA_STACK_TEST_RECORDING_DIR="tests/integration/recordings"
+
+          if [ "$SHOULD_RECORD" == "true" ]; then
+            export LLAMA_STACK_TEST_INFERENCE_MODE="record"
+          else
+            export LLAMA_STACK_TEST_INFERENCE_MODE="replay"
+          fi
+
           if [ "${{ matrix.provider }}" == "ollama" ]; then
             export OLLAMA_URL="http://0.0.0.0:11434"
             export TEXT_MODEL=ollama/llama3.2:3b-instruct-fp16
@@ -129,7 +170,6 @@ jobs:
             EXCLUDE_TESTS="${EXCLUDE_TESTS} or test_inference_store_tool_calls"
           fi
 
-
           uv run pytest -s -v tests/integration/${{ matrix.test-type }} --stack-config=${stack_config} \
             -k "not( ${EXCLUDE_TESTS} )" \
             --text-model=$TEXT_MODEL \
@@ -137,6 +177,20 @@ jobs:
             --color=yes ${EXTRA_PARAMS} \
             --capture=tee-sys | tee pytest-${{ matrix.test-type }}.log
 
+      - name: Update the PR if tests/integration/recordings/ has changed
+        if: ${{ env.SHOULD_RECORD == 'true' }}
+        run: |
+          if ! git diff --quiet tests/integration/recordings/; then
+            echo "Updating PR with updated recordings"
+            git config --local user.email "github-actions[bot]@users.noreply.github.com"
+            git config --local user.name "github-actions[bot]"
+            git add tests/integration/recordings/
+            git commit -m "Update recordings from integration tests"
+            git push origin HEAD:${{ github.head_ref }}
+          else
+            echo "No changes to recordings detected"
+          fi
+
       - name: Check Storage and Memory Available After Tests
         if: ${{ always() }}
         run: |
@@ -144,13 +198,13 @@ jobs:
           df -h
 
       - name: Write inference logs to file
-        if: ${{ always() }}
+        if: ${{ env.SHOULD_RECORD == 'true' }}
         run: |
           sudo docker logs ollama > ollama.log || true
           sudo docker logs vllm > vllm.log || true
 
       - name: Upload all logs to artifacts
-        if: ${{ always() }}
+        if: ${{ env.SHOULD_RECORD == 'true' }}
         uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
           name: logs-${{ github.run_id }}-${{ github.run_attempt }}-${{ matrix.provider }}-${{ matrix.client-type }}-${{ matrix.test-type }}-${{ matrix.python-version }}-${{ matrix.client-version }}

tests/integration/recordings/responses/0ff78129bb3a.json

@@ -0,0 +1,167 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\nReturns the boiling point of a liquid in Celsius or Fahrenheit.\n\n:param liquid_name: The name of the liquid\n:param celsius: Whether to return the boiling point in Celsius\n:return: The boiling point of the liquid in Celcius or Fahrenheit\n\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n[greet_everyone(url=\"world\")]<|eot_id|><|start_header_id|>ipython<|end_header_id|>\n\nHello, world!<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.143606Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "How",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.186151Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " can",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.229036Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " I",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.271516Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " assist",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.316272Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " you",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.361005Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " further",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.404689Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "?",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.447699Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 456939083,
+          "load_duration": 79653292,
+          "prompt_eval_count": 471,
+          "prompt_eval_duration": 71724667,
+          "eval_count": 8,
+          "eval_duration": 304859000,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/4597743bcd2a.json

@@ -0,0 +1,185 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\nReturns the boiling point of a liquid in Celsius or Fahrenheit.\n\n:param liquid_name: The name of the liquid\n:param celsius: Whether to return the boiling point in Celsius\n:return: The boiling point of the liquid in Celcius or Fahrenheit\n\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.476678Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "[g",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.520346Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "reet",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.563375Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_every",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.606256Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "one",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.649215Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "(url",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.692049Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "=\"",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.734316Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "world",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.776615Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "\")]",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:17.819266Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 5629478417,
+          "load_duration": 4092162625,
+          "prompt_eval_count": 448,
+          "prompt_eval_duration": 1191158583,
+          "eval_count": 9,
+          "eval_duration": 343915792,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/4a3a4447b16b.json

@@ -14,7 +14,7 @@
         "models": [
           {
             "model": "nomic-embed-text:latest",
-            "modified_at": "2025-07-29T14:55:45.755906-07:00",
+            "modified_at": "2025-07-29T16:46:26.304701-07:00",
             "digest": "0a109f422b47e3a30ba2b10eca18548e944e8a23073ee3f3e947efcf3c45e59f",
             "size": 274302450,
             "details": {

tests/integration/recordings/responses/9c28ec9ac338.json

@@ -0,0 +1,347 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\n        Returns the boiling point of a liquid in Celsius or Fahrenheit.\n\n        :param liquid_name: The name of the liquid\n        :param celsius: Whether to return the boiling point in Celsius\n        :return: The boiling point of the liquid in Celcius or Fahrenheit\n        \",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n[greet_everyone(url=\"world\")]<|eot_id|><|start_header_id|>ipython<|end_header_id|>\n\nHello, world!<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\nHow can I assist you further?<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nWhat is the boiling point of polyjuice? Use tools to answer.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.316207Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "[",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.358611Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "get",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.401272Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_bo",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.444321Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "iling",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.48795Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_point",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.530158Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "(",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.573318Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "liquid",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.616297Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_name",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.659527Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "='",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.702422Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "poly",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.745894Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "ju",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.788811Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "ice",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.831618Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "',",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.874469Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " c",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.917372Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "elsius",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.960558Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "=True",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:37.004223Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": ")]",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:37.046563Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 845522667,
+          "load_duration": 47784875,
+          "prompt_eval_count": 511,
+          "prompt_eval_duration": 66135292,
+          "eval_count": 18,
+          "eval_duration": 730999291,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/bd356b27a085.json

@@ -0,0 +1,167 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\n        Returns the boiling point of a liquid in Celsius or Fahrenheit.\n\n        :param liquid_name: The name of the liquid\n        :param celsius: Whether to return the boiling point in Celsius\n        :return: The boiling point of the liquid in Celcius or Fahrenheit\n        \",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n[greet_everyone(url=\"world\")]<|eot_id|><|start_header_id|>ipython<|end_header_id|>\n\nHello, world!<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.850399Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "How",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.89419Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " can",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.938049Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " I",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.980392Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " assist",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.023004Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " you",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.065467Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " further",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.108189Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "?",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:36.150902Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 468910417,
+          "load_duration": 93969000,
+          "prompt_eval_count": 479,
+          "prompt_eval_duration": 72596750,
+          "eval_count": 8,
+          "eval_duration": 301590375,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/c7582fa7c2c4.json

@@ -0,0 +1,347 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\nReturns the boiling point of a liquid in Celsius or Fahrenheit.\n\n:param liquid_name: The name of the liquid\n:param celsius: Whether to return the boiling point in Celsius\n:return: The boiling point of the liquid in Celcius or Fahrenheit\n\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n[greet_everyone(url=\"world\")]<|eot_id|><|start_header_id|>ipython<|end_header_id|>\n\nHello, world!<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\nHow can I assist you further?<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nWhat is the boiling point of polyjuice? Use tools to answer.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.64197Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "[",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.687885Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "get",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.73112Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_bo",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.774191Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "iling",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.816695Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_point",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.859121Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "(",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.901585Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "liquid",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.943788Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_name",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:18.986429Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "='",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.029894Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "poly",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.073113Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "ju",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.116671Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "ice",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.159456Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "',",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.203354Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": " c",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.246192Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "elsius",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.290499Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "=True",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.334562Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": ")]",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:26:19.380415Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 881889250,
+          "load_duration": 69966916,
+          "prompt_eval_count": 503,
+          "prompt_eval_duration": 70368167,
+          "eval_count": 18,
+          "eval_duration": 740885458,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/ef757a75ed08.json

@@ -0,0 +1,185 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:3b-instruct-fp16",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\n        Returns the boiling point of a liquid in Celsius or Fahrenheit.\n\n        :param liquid_name: The name of the liquid\n        :param celsius: Whether to return the boiling point in Celsius\n        :return: The boiling point of the liquid in Celcius or Fahrenheit\n        \",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:3b-instruct-fp16"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.212563Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "[g",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.254896Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "reet",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.297152Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "_every",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.339477Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "one",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.382245Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "(url",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.423387Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "=\"",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.465286Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "world",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.507249Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "\")]",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:3b-instruct-fp16",
+          "created_at": "2025-07-29T23:46:35.549072Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 5519843458,
+          "load_duration": 4110366375,
+          "prompt_eval_count": 456,
+          "prompt_eval_duration": 1070783708,
+          "eval_count": 9,
+          "eval_duration": 337120750,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/recordings/responses/f3c3afbd9b7e.json

@@ -0,0 +1,59 @@
+{
+  "request": {
+    "method": "POST",
+    "url": "http://localhost:11434/api/generate",
+    "headers": {},
+    "body": {
+      "model": "llama3.2:1b",
+      "raw": true,
+      "prompt": "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a helpful assistant. You have access to functions, but you should only use them if they are required.\nYou are an expert in composing functions. You are given a question and a set of possible functions.\nBased on the question, you may or may not need to make one function/tool call to achieve the purpose.\n\nIf you decide to invoke any of the function(s), you MUST put it in the format of [func_name1(params_name1=params_value1, params_name2=params_value2...), func_name2(params)]\nIf you decide to invoke a function, you SHOULD NOT include any other text in the response. besides the function call in the above format.\nFor a boolean parameter, be sure to use `True` or `False` (capitalized) for the value.\n\n\nHere is a list of functions in JSON format that you can invoke.\n\n[\n    {\n        \"name\": \"greet_everyone\",\n        \"description\": \"\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"url\"],\n            \"properties\": {\n                \"url\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    },\n    {\n        \"name\": \"get_boiling_point\",\n        \"description\": \"\nReturns the boiling point of a liquid in Celsius or Fahrenheit.\n\n:param liquid_name: The name of the liquid\n:param celsius: Whether to return the boiling point in Celsius\n:return: The boiling point of the liquid in Celcius or Fahrenheit\n\",\n        \"parameters\": {\n            \"type\": \"dict\",\n            \"required\": [\"liquid_name\", \"celsius\"],\n            \"properties\": {\n                \"liquid_name\": {\n                    \"type\": \"string\",\n                    \"description\": \"\"\n                },\n                \"celsius\": {\n                    \"type\": \"boolean\",\n                    \"description\": \"\"\n                }\n            }\n        }\n    }\n]\n\nYou can answer general questions or invoke tools when necessary.\nIn addition to tool calls, you should also augment your responses by using the tool outputs.\nYou are a helpful assistant.<|eot_id|><|start_header_id|>user<|end_header_id|>\n\nSay hi to the world. Use tools to do so.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n",
+      "options": {
+        "temperature": 0.0
+      },
+      "stream": true
+    },
+    "endpoint": "/api/generate",
+    "model": "llama3.2:1b"
+  },
+  "response": {
+    "body": [
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:1b",
+          "created_at": "2025-07-29T23:23:09.553247Z",
+          "done": false,
+          "done_reason": null,
+          "total_duration": null,
+          "load_duration": null,
+          "prompt_eval_count": null,
+          "prompt_eval_duration": null,
+          "eval_count": null,
+          "eval_duration": null,
+          "response": "Hi",
+          "thinking": null,
+          "context": null
+        }
+      },
+      {
+        "__type__": "ollama._types.GenerateResponse",
+        "__data__": {
+          "model": "llama3.2:1b",
+          "created_at": "2025-07-29T23:23:09.564069Z",
+          "done": true,
+          "done_reason": "stop",
+          "total_duration": 2125493250,
+          "load_duration": 1610279708,
+          "prompt_eval_count": 448,
+          "prompt_eval_duration": 502413125,
+          "eval_count": 2,
+          "eval_duration": 11573709,
+          "response": "",
+          "thinking": null,
+          "context": null
+        }
+      }
+    ],
+    "is_streaming": true
+  }
+}

tests/integration/tool_runtime/test_mcp.py

@@ -10,7 +10,6 @@ import pytest
 from llama_stack_client import Agent
 
 from llama_stack import LlamaStackAsLibraryClient
-from llama_stack.apis.models import ModelType
 from llama_stack.distribution.datatypes import AuthenticationRequiredError
 
 AUTH_TOKEN = "test-token"
@@ -24,7 +23,7 @@ def mcp_server():
         yield mcp_server_info
 
 
-def test_mcp_invocation(llama_stack_client, mcp_server):
+def test_mcp_invocation(llama_stack_client, text_model_id, mcp_server):
     if not isinstance(llama_stack_client, LlamaStackAsLibraryClient):
         pytest.skip("The local MCP server only reliably reachable from library client.")
 
@@ -69,14 +68,10 @@ def test_mcp_invocation(llama_stack_client, mcp_server):
     assert content[0].type == "text"
     assert content[0].text == "Hello, world!"
 
-    models = [
+    print(f"Using model: {text_model_id}")
-        m for m in llama_stack_client.models.list() if m.model_type == ModelType.llm and "guard" not in m.identifier
-    ]
-    model_id = models[0].identifier
-    print(f"Using model: {model_id}")
     agent = Agent(
         client=llama_stack_client,
-        model=model_id,
+        model=text_model_id,
         instructions="You are a helpful assistant.",
         tools=[test_toolgroup_id],
     )