Feats: Added google / Gemini provider

features/google_provider.feature

@@ -0,0 +1,38 @@
+Feature: Google provider adapter coverage
+  As a maintainer validating real provider adapters
+  I want Behave scenarios for the Google chat provider
+  So that constructor guards and LangChain handoffs stay regression tested
+
+  @unit @providers @google
+  Scenario: Google provider instantiates ChatGoogleGenerativeAI with provided credentials
+    Given I have sample provider domain inputs
+    When I create a Google chat provider with API key "sk-google-unit" and model "gemini-2.0-flash"
+    And I request plan generation from the Google provider
+    Then the Google provider should construct ChatGoogleGenerativeAI with api key "sk-google-unit" and model "gemini-2.0-flash"
+    And the Google provider metadata should report name "google" and model "gemini-2.0-flash"
+
+  @unit @providers @google
+  Scenario: Google provider forwards extra kwargs to ChatGoogleGenerativeAI
+    Given I have sample provider domain inputs
+    And I set the Google provider extra kwargs "temperature=0.25,max_output_tokens=2048"
+    When I create a Google chat provider with API key "sk-google-unit" and model "gemini-2.0-pro"
+    And I request plan generation from the Google provider
+    Then the Google provider should construct ChatGoogleGenerativeAI with api key "sk-google-unit" and model "gemini-2.0-pro"
+    And the Google provider should include kwargs "temperature=0.25,max_output_tokens=2048" in the ChatGoogleGenerativeAI call
+    And the Google provider metadata should report name "google" and model "gemini-2.0-pro"
+
+  @unit @providers @google
+  Scenario: Google provider returns generated changes with token count
+    Given I have sample provider domain inputs
+    And the plan generation graph returns a generated change for "app/google_provider.py"
+    And the Google provider token estimator returns 256 tokens
+    When I create a Google chat provider with API key "sk-google-unit" and model "gemini-2.0-flash"
+    And I request plan generation from the Google provider
+    Then the Google provider response should include 1 generated change for "app/google_provider.py"
+    And the Google provider response token count should equal 256
+
+  @unit @providers @google
+  Scenario: Google provider rejects missing API key
+    Given I have sample provider domain inputs
+    When I attempt to create a Google chat provider without an API key
+    Then the Google provider creation should fail with error "Google API key is required"

features/provider_registry_coverage.feature

@@ -359,6 +359,14 @@ Feature: Provider Registry Coverage
     And the stubbed OpenRouter client headers should include "HTTP-Referer" value "ExampleApp"
     And the stubbed OpenRouter client headers should include "X-Title" value "ExampleApp"
 
+  @unit @providers @registry
+  Scenario: _create_provider_llm coerces OpenRouter default header mappings to strings
+    Given I have a ProviderRegistry with "openrouter" API key set
+    And the "OpenRouter" LangChain client is stubbed
+    And the OpenRouter default headers include numeric entries
+    When I call _create_provider_llm for provider "openrouter" with model "router-coerce" and custom default headers
+    Then the stubbed OpenRouter client headers should include "123" value "456"
+
   @unit @providers @registry
   Scenario: _create_provider_llm raises error for unsupported provider
     Given I have a ProviderRegistry instance
@@ -379,6 +387,13 @@ Feature: Provider Registry Coverage
     Then a provider registry ValueError should be raised
     And the provider registry error should mention unknown provider type
 
+  @unit @providers @registry
+  Scenario: Create AI provider raises error when Google API key is missing
+    Given I have a ProviderRegistry with no API keys
+    When I try to create an AI provider for provider "google"
+    Then a provider registry ValueError should be raised
+    And the provider registry error should mention missing "GOOGLE_API_KEY" environment variable
+
   @unit @providers @registry
   Scenario: Create AI provider succeeds for configured provider string
     Given I have a ProviderRegistry with OpenAI API key set
@@ -388,6 +403,15 @@ Feature: Provider Registry Coverage
     And the provider registry AI provider name should be "openai"
     And the provider registry AI provider model_id should be "gpt-4o"
 
+  @unit @providers @registry
+  Scenario: Create AI provider returns Google adapter
+    Given I have a ProviderRegistry with "google" API key set
+    And CLEVERAGENTS_DEFAULT_MODEL is not set
+    When I create an AI provider for provider "google"
+    Then the provider registry AI provider should be a GoogleChatProvider
+    And the provider registry AI provider name should be "google"
+    And the provider registry AI provider model_id should be "gemini-2.0-flash"
+
   @unit @providers @registry
   Scenario: Create AI provider without specifying provider uses fallback
     Given I have a ProviderRegistry with OpenAI API key set
@@ -397,6 +421,7 @@ Feature: Provider Registry Coverage
     Then the provider registry AI provider should be a LangChainChatProvider
     And the provider registry AI provider name should be "openai"
 
+
   @unit @providers @registry
   Scenario: Create AI provider honors explicit model overrides
     Given I have a ProviderRegistry with OpenAI API key set

features/steps/google_provider_steps.py

@@ -0,0 +1,208 @@
+from __future__ import annotations
+
+import ast
+from typing import Any
+from unittest.mock import MagicMock, patch
+
+from behave import given, then, when
+
+from cleveragents.providers.llm.google_provider import GoogleChatProvider
+
+
+def _register_cleanup(context, cleanup):
+    if hasattr(context, "add_cleanup"):
+        context.add_cleanup(cleanup)
+    else:
+        cleanup_handlers = getattr(context, "_cleanup_handlers", [])
+        cleanup_handlers.append(cleanup)
+        context._cleanup_handlers = cleanup_handlers
+
+
+def _parse_kwargs_string(kwargs_string: str) -> dict[str, Any]:
+    if not kwargs_string:
+        return {}
+    parsed: dict[str, Any] = {}
+    for entry in kwargs_string.split(","):
+        entry = entry.strip()
+        if not entry:
+            continue
+        if "=" not in entry:
+            parsed[entry] = True
+            continue
+        key, value = entry.split("=", 1)
+        key = key.strip()
+        value = value.strip()
+        try:
+            parsed_value = ast.literal_eval(value)
+        except Exception:
+            parsed_value = value
+        parsed[key] = parsed_value
+    return parsed
+
+
+def _setup_plan_generation_graph(context) -> MagicMock:
+    patcher = patch(
+        "cleveragents.providers.llm.langchain_chat_provider.PlanGenerationGraph"
+    )
+    context.plan_generation_patcher = patcher
+    mock_graph_class = patcher.start()
+    _register_cleanup(context, patcher.stop)
+    mock_graph_instance = MagicMock(name="PlanGenerationGraphInstance")
+
+    default_state = {
+        "generated_changes": [],
+        "validation_result": {"status": "PASS"},
+        "error": None,
+    }
+    state_override = getattr(context, "plan_generation_state_override", None)
+    mock_graph_instance.invoke.return_value = state_override or default_state
+
+    invoke_side_effect = getattr(context, "plan_generation_invoke_side_effect", None)
+    if invoke_side_effect is not None:
+        mock_graph_instance.invoke.side_effect = invoke_side_effect
+
+    stream_events = getattr(context, "plan_generation_stream_events", None)
+    if stream_events is None:
+        mock_graph_instance.stream.return_value = iter(())
+    else:
+        events_copy = list(stream_events)
+
+        def _stream(*_args, **_kwargs):
+            for event in events_copy:
+                yield event
+
+        mock_graph_instance.stream.side_effect = _stream
+
+    mock_graph_class.return_value = mock_graph_instance
+    context.plan_generation_graph = mock_graph_instance
+    context.plan_generation_graph_class = mock_graph_class
+    return mock_graph_instance
+
+
+@given('I set the Google provider extra kwargs "{kwargs_string}"')
+def step_set_google_provider_kwargs(context, kwargs_string):
+    context.google_provider_kwargs = _parse_kwargs_string(kwargs_string)
+
+
+@given("the Google provider token estimator returns {token_count:d} tokens")
+def step_set_google_token_estimator(context, token_count):
+    context.google_token_count = token_count
+
+
+@when('I create a Google chat provider with API key "{api_key}" and model "{model_id}"')
+def step_create_google_provider(context, api_key, model_id):
+    patcher = patch("cleveragents.providers.llm.google_provider.ChatGoogleGenerativeAI")
+    context.chat_google_patcher = patcher
+    mock_chat_class = patcher.start()
+    _register_cleanup(context, patcher.stop)
+    mock_chat_instance = MagicMock(name="ChatGoogleGenerativeAIInstance")
+    mock_chat_instance.get_num_tokens = MagicMock(return_value=0)
+    token_override = getattr(context, "google_token_count", None)
+    if token_override is not None:
+        mock_chat_instance.get_num_tokens.return_value = token_override
+    mock_chat_class.return_value = mock_chat_instance
+
+    extra_kwargs = dict(getattr(context, "google_provider_kwargs", {}))
+
+    context.provider = GoogleChatProvider(
+        api_key=api_key,
+        model=model_id,
+        **extra_kwargs,
+    )
+    context.chat_google_class = mock_chat_class
+    context.chat_google_instance = mock_chat_instance
+
+
+@when("I attempt to create a Google chat provider without an API key")
+def step_google_provider_without_api_key(context):
+    context.google_provider_error = None
+    try:
+        GoogleChatProvider(api_key="", model="gemini-2.0-flash")
+    except Exception as exc:  # pragma: no cover - defensive
+        context.google_provider_error = exc
+
+
+@when("I request plan generation from the Google provider")
+def step_request_google_plan_generation(context):
+    _setup_plan_generation_graph(context)
+    context.response = context.provider.generate_changes(
+        context.project,
+        context.plan,
+        context.contexts,
+    )
+    context.chat_google_call = context.chat_google_class.call_args
+    context.plan_generation_graph_call = context.plan_generation_graph_class.call_args
+
+
+@then(
+    'the Google provider should construct ChatGoogleGenerativeAI with api key "{api_key}" and model "{model_id}"'
+)
+def step_assert_google_constructor(context, api_key, model_id):
+    call = getattr(context, "chat_google_call", None)
+    assert call is not None, "ChatGoogleGenerativeAI should have been called"
+    call_args, call_kwargs = call
+    assert call_args == (), "ChatGoogleGenerativeAI should use keyword arguments"
+    assert call_kwargs["api_key"] == api_key
+    assert call_kwargs["model"] == model_id
+
+    graph_call = getattr(context, "plan_generation_graph_call", None)
+    assert graph_call is not None, "PlanGenerationGraph should receive the LLM"
+    _, graph_kwargs = graph_call
+    assert graph_kwargs["llm"] is context.chat_google_instance
+
+
+@then(
+    'the Google provider should include kwargs "{kwargs_string}" in the ChatGoogleGenerativeAI call'
+)
+def step_assert_google_kwargs(context, kwargs_string):
+    call = getattr(context, "chat_google_call", None)
+    assert call is not None, "ChatGoogleGenerativeAI should have been called"
+    _, call_kwargs = call
+    expected = _parse_kwargs_string(kwargs_string)
+    for key, value in expected.items():
+        assert key in call_kwargs, f"Expected {key} in ChatGoogleGenerativeAI kwargs"
+        assert call_kwargs[key] == value, (
+            f"Expected ChatGoogleGenerativeAI kwargs[{key!r}] to equal {value!r}"
+        )
+
+
+@then(
+    'the Google provider metadata should report name "{expected_name}" and model "{expected_model}"'
+)
+def step_assert_google_metadata(context, expected_name, expected_model):
+    provider = getattr(context, "provider", None)
+    assert provider is not None, "Google provider should exist"
+    assert provider.name == expected_name
+    assert provider.model_id == expected_model
+    response = getattr(context, "response", None)
+    if response is not None:
+        assert response.model_used == expected_model
+
+
+@then(
+    'the Google provider response should include {count:d} generated change for "{file_path}"'
+)
+def step_assert_google_generated_change(context, count, file_path):
+    response = getattr(context, "response", None)
+    assert response is not None, "Provider response should exist"
+    assert len(response.changes) == count, (
+        f"Expected {count} changes, got {len(response.changes)}"
+    )
+    assert any(
+        getattr(change, "file_path", None) == file_path for change in response.changes
+    ), f"Expected change for {file_path}"
+
+
+@then("the Google provider response token count should equal {expected:d}")
+def step_assert_google_token_count(context, expected):
+    response = getattr(context, "response", None)
+    assert response is not None
+    assert response.token_count == expected
+
+
+@then('the Google provider creation should fail with error "{message}"')
+def step_assert_google_creation_error(context, message):
+    error = getattr(context, "google_provider_error", None)
+    assert error is not None, "Expected the provider to raise an error"
+    assert isinstance(error, ValueError)
+    assert str(error) == message

features/steps/openai_provider_steps.py

@@ -5,14 +5,13 @@ from typing import Any
 from unittest.mock import MagicMock, patch
 
 from behave import given, then, when
-
 from tenacity import RetryError, Retrying, stop_after_attempt
 
 from cleveragents.domain.models.core import (
     Context,
+    OperationType,
     Plan,
     Project,
-    OperationType,
 )
 from cleveragents.providers.llm.openai_provider import OpenAIChatProvider
 

features/steps/provider_registry_steps.py

@@ -9,6 +9,7 @@ from unittest.mock import MagicMock
 
 from behave import given, then, when  # type: ignore[import-untyped]
 
+from cleveragents.providers.llm.google_provider import GoogleChatProvider
 from cleveragents.providers.llm.langchain_chat_provider import LangChainChatProvider
 from cleveragents.providers.registry import (
     ProviderCapabilities,
@@ -262,6 +263,11 @@ def step_impl_set_openrouter_org(context: Any, value: str) -> None:
     context.registry._settings.openrouter_organization = value
 
 
+@given("the OpenRouter default headers include numeric entries")
+def step_impl_set_openrouter_default_headers(context: Any) -> None:
+    context.openrouter_default_headers = {123: 456, "X-Debug": 789}
+
+
 @given("the Azure endpoint setting is cleared")
 def step_impl_clear_azure_endpoint(context: Any) -> None:
     assert hasattr(context, "registry"), "Registry must exist before clearing endpoint"
@@ -496,6 +502,21 @@ def step_impl_call_private_llm(context: Any, value: str, model: str) -> None:
     context.created_llm = context.registry._create_provider_llm(provider_type, model)
 
 
+@when(
+    'I call _create_provider_llm for provider "{value}" with model "{model}" and custom default headers'
+)
+def step_impl_call_private_llm_with_headers(
+    context: Any, value: str, model: str
+) -> None:
+    provider_enum = _ensure_provider_type(context)
+    provider_type = getattr(provider_enum, value.upper())
+    default_headers = getattr(context, "openrouter_default_headers", None)
+    assert default_headers is not None, "OpenRouter default headers were not configured"
+    context.created_llm = context.registry._create_provider_llm(
+        provider_type, model, default_headers=default_headers
+    )
+
+
 @when('I try to call _create_provider_llm for provider "{value}" with model "{model}"')
 def step_impl_try_private_llm(context: Any, value: str, model: str) -> None:
     provider_enum = _ensure_provider_type(context)
@@ -743,11 +764,27 @@ def step_impl_error_mentions_missing_endpoint(context: Any) -> None:
     assert "azure" in message and "endpoint" in message
 
 
+@then(
+    'the provider registry error should mention missing "{env_var}" environment variable'
+)
+def step_impl_error_mentions_missing_env(context: Any, env_var: str) -> None:
+    assert context.error is not None
+    message = str(context.error)
+    assert env_var in message, (
+        f"Expected error message to mention {env_var!r}, got {message!r}"
+    )
+
+
 @then("the provider registry AI provider should be a LangChainChatProvider")
 def step_impl_ai_provider_is_langchain(context: Any) -> None:
     assert isinstance(context.ai_provider, LangChainChatProvider)
 
 
+@then("the provider registry AI provider should be a GoogleChatProvider")
+def step_impl_ai_provider_is_google(context: Any) -> None:
+    assert isinstance(context.ai_provider, GoogleChatProvider)
+
+
 @then('the provider registry AI provider name should be "{value}"')
 def step_impl_ai_provider_name(context: Any, value: str) -> None:
     assert context.ai_provider.name == value

implementation_plan.md

@@ -1693,6 +1693,12 @@ Notes: Record provider-specific nuances, API limitations, and testing fixtures.
 - Authored `features/anthropic_provider.feature:1-19` with matching steps in `features/steps/anthropic_provider_steps.py:1-129` to cover missing-key errors, constructor overrides, and metadata reporting while asserting `PlanGenerationGraph` receives the mocked `ChatAnthropic` instance.
 - Executed `behave features/anthropic_provider.feature` locally to confirm both Anthropic scenarios pass alongside the refreshed OpenAI suite.
 
+**2025-12-09: Google provider adapter + registry coverage**
+- Added `src/cleveragents/providers/llm/google_provider.py:1-40` implementing `GoogleChatProvider`, including API key validation, overridable kwargs, and a LangChain `ChatGoogleGenerativeAI` factory wired for streaming.
+- Updated `src/cleveragents/providers/registry.py:585-614` so `create_ai_provider` now instantiates `GoogleChatProvider` when Google credentials are configured, preserving capability metadata for other providers.
+- Authored `features/google_provider.feature:1-39` with steps in `features/steps/google_provider_steps.py:1-209` to cover constructor guards, kwarg forwarding, token accounting, and plan-generation graph handoff; expanded `features/provider_registry_coverage.feature:390-422` plus `features/steps/provider_registry_steps.py:12-555` to assert the registry returns the Google adapter instance.
+- Added Robot coverage via `robot/google_provider.robot:1-169` for generate/stream/error/retry paths, and verified everything with `nox -s unit_tests -- features/google_provider.feature features/provider_registry_coverage.feature` followed by `nox -s integration_tests -- robot/google_provider.robot` (both passing).
+
 **Additional LangChain/LangGraph Tasks for Phase 4:**
 - Implement `CodeGenerationGraph` using LangGraph for multi-step code generation
 - Create `AutoDebugGraph` with conditional edges for error detection and retry
@@ -4198,7 +4204,7 @@ If you can do all of the above by end of Day 1, you're on track!
         - [X] Replace the placeholder `generate_changes()` with the real LangChain-powered workflow by invoking `LangChainChatProvider.generate_changes`, ensuring PlanGenerationGraph produces structured plan deltas and token estimates.
         - [X] Add streaming support that forwards `ChatOpenAI.stream()` events to PlanService/CLI consumers, including progress callbacks and graceful cancellation semantics (implemented via `LangChainChatProvider.stream_changes()` and consumed by `PlanService.generate_plan_streaming`).
         - [X] Capture token usage + cost metrics from the `LLMResult` and persist them on `ProviderResponse.token_count`, logging through structlog with LangSmith tags (see `_resolve_token_count/_log_usage` + new plan-service persistence path).
-        - [ ] Expand Behave + Robot coverage to exercise real OpenAI provider flows (using patched clients) covering success, streaming, error, and retry scenarios.
+        - [X] Expand Behave + Robot coverage to exercise real OpenAI provider flows (using patched clients) covering success, streaming, error, and retry scenarios.
         - [X] Implement Anthropic chat provider adapter (Claude) with parity tests (`src/cleveragents/providers/llm/anthropic_provider.py:1-36`).
         - [ ] Implement Google Gemini provider adapter with parity tests.
         - [ ] Implement OpenRouter provider adapter with org header support.

robot/google_provider.robot

@@ -0,0 +1,168 @@
+*** Settings ***
+Resource          common.resource
+Library           Process
+Library           OperatingSystem
+Library           Collections
+
+*** Variables ***
+${PYTHON}         python
+${SRC_DIR}        ${CURDIR}/..
+
+*** Test Cases ***
+Google Provider Generates Changes
+    ${script}=    Catenate    SEPARATOR=\n
+    ...    import sys
+    ...    from unittest.mock import MagicMock, patch
+    ...    sys.path.insert(0, '${SRC_DIR}/src')
+    ...    from cleveragents.providers.llm.google_provider import GoogleChatProvider
+    ...    project = MagicMock(name='Project')
+    ...    plan = MagicMock(name='Plan')
+    ...    plan.prompt = 'Add logging'
+    ...    contexts = [MagicMock(name='Context')]
+    ...    contexts[0].content = 'ctx'
+    ...    chat_patcher = patch('cleveragents.providers.llm.google_provider.ChatGoogleGenerativeAI')
+    ...    graph_patcher = patch('cleveragents.providers.llm.langchain_chat_provider.PlanGenerationGraph')
+    ...    chat_cls = chat_patcher.start()
+    ...    graph_cls = graph_patcher.start()
+    ...    chat_instance = MagicMock(name='ChatGoogleGenerativeAIInstance')
+    ...    chat_instance.get_num_tokens.return_value = 64
+    ...    chat_cls.return_value = chat_instance
+    ...    change = {'plan_id': 1, 'file_path': 'robot_google.py', 'operation': 'modify', 'new_content': '# code'}
+    ...    graph_instance = MagicMock(name='GraphInstance')
+    ...    graph_instance.invoke.return_value = {'generated_changes': [change], 'validation_result': {'status': 'PASS'}, 'error': None}
+    ...    graph_cls.return_value = graph_instance
+    ...    provider = GoogleChatProvider(api_key='sk-google-robot', model='gemini-2.0-flash')
+    ...    response = provider.generate_changes(project, plan, contexts)
+    ...    chat_patcher.stop()
+    ...    graph_patcher.stop()
+    ...    assert len(response.changes) == 1
+    ...    assert response.changes[0].file_path == 'robot_google.py'
+    ...    assert response.token_count == 64
+    ...    print('google-provider-success')
+     ${result}=    Run Process    ${PYTHON}    -c    ${script}
+     Log Process Failure    ${result}
+
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    google-provider-success
+
+Google Provider Streams Events
+    ${script}=    Catenate    SEPARATOR=\n
+    ...    import sys
+    ...    from unittest.mock import MagicMock, patch
+    ...    sys.path.insert(0, '${SRC_DIR}/src')
+    ...    from cleveragents.providers.llm.google_provider import GoogleChatProvider
+    ...    project = MagicMock(name='Project')
+    ...    plan = MagicMock(name='Plan')
+    ...    plan.prompt = 'Stream plan'
+    ...    contexts = [MagicMock(name='Context')]
+    ...    contexts[0].content = 'ctx'
+    ...    chat_patcher = patch('cleveragents.providers.llm.google_provider.ChatGoogleGenerativeAI')
+    ...    graph_patcher = patch('cleveragents.providers.llm.langchain_chat_provider.PlanGenerationGraph')
+    ...    chat_cls = chat_patcher.start()
+    ...    graph_cls = graph_patcher.start()
+    ...    chat_instance = MagicMock(name='ChatGoogleGenerativeAIInstance')
+    ...    chat_instance.get_num_tokens.return_value = 32
+    ...    chat_cls.return_value = chat_instance
+    ...    graph_instance = MagicMock(name='GraphInstance')
+    ...    events = [
+    ...        {'load_context': {'status': 'ok'}},
+    ...        {'generate_plan': {'generated_changes': [{'plan_id': 1, 'file_path': 'stream_google.py', 'operation': 'modify', 'new_content': '# stream'}]}},
+    ...        {'validate': {'validation_result': {'status': 'PASS'}}},
+    ...    ]
+    ...    graph_instance.stream.side_effect = lambda *_args, **_kwargs: iter(events)
+    ...    graph_instance.invoke.return_value = {'generated_changes': [], 'validation_result': {'status': 'PASS'}, 'error': None}
+    ...    graph_cls.return_value = graph_instance
+    ...    provider = GoogleChatProvider(api_key='sk-google-robot', model='gemini-2.0-flash')
+    ...    streamed = list(provider.stream_changes(project, plan, contexts))
+    ...    chat_patcher.stop()
+    ...    graph_patcher.stop()
+    ...    assert streamed[-1]['__end__']['response'].model_used == 'gemini-2.0-flash'
+    ...    assert len(streamed[-1]['__end__']['response'].changes) == 1
+    ...    print('google-stream-success')
+     ${result}=    Run Process    ${PYTHON}    -c    ${script}
+     Log Process Failure    ${result}
+
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    google-stream-success
+
+Google Provider Surfaces Graph Errors
+    ${script}=    Catenate    SEPARATOR=\n
+    ...    import sys
+    ...    from unittest.mock import MagicMock, patch
+    ...    sys.path.insert(0, '${SRC_DIR}/src')
+    ...    from cleveragents.providers.llm.google_provider import GoogleChatProvider
+    ...    project = MagicMock(name='Project')
+    ...    plan = MagicMock(name='Plan')
+    ...    plan.prompt = 'Error plan'
+    ...    contexts = [MagicMock(name='Context')]
+    ...    contexts[0].content = 'ctx'
+    ...    chat_patcher = patch('cleveragents.providers.llm.google_provider.ChatGoogleGenerativeAI')
+    ...    graph_patcher = patch('cleveragents.providers.llm.langchain_chat_provider.PlanGenerationGraph')
+    ...    chat_cls = chat_patcher.start()
+    ...    graph_cls = graph_patcher.start()
+    ...    chat_instance = MagicMock(name='ChatGoogleGenerativeAIInstance')
+    ...    chat_instance.get_num_tokens.return_value = 0
+    ...    chat_cls.return_value = chat_instance
+    ...    graph_instance = MagicMock(name='GraphInstance')
+    ...    graph_instance.invoke.side_effect = ValueError('graph exploded')
+    ...    graph_cls.return_value = graph_instance
+    ...    provider = GoogleChatProvider(api_key='sk-google-robot', model='gemini-2.0-flash')
+    ...    response = provider.generate_changes(project, plan, contexts)
+    ...    chat_patcher.stop()
+    ...    graph_patcher.stop()
+    ...    assert response.error_message == 'graph exploded'
+    ...    assert response.changes == []
+    ...    print('google-provider-error')
+     ${result}=    Run Process    ${PYTHON}    -c    ${script}
+     Log Process Failure    ${result}
+
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    google-provider-error
+
+Google Provider Reports Retry Exhaustion
+    ${script}=    Catenate    SEPARATOR=\n
+    ...    import sys
+    ...    from unittest.mock import MagicMock, patch
+    ...    from tenacity import RetryError, Future
+    ...    sys.path.insert(0, '${SRC_DIR}/src')
+    ...    from cleveragents.providers.llm.google_provider import GoogleChatProvider
+    ...    project = MagicMock(name='Project')
+    ...    plan = MagicMock(name='Plan')
+    ...    plan.prompt = 'Retry plan'
+    ...    contexts = [MagicMock(name='Context')]
+    ...    contexts[0].content = 'ctx'
+    ...    future = Future(1)
+    ...    future.set_exception(RuntimeError('deadline reached'))
+    ...    retry_error = RetryError(future)
+    ...    chat_patcher = patch('cleveragents.providers.llm.google_provider.ChatGoogleGenerativeAI')
+    ...    graph_patcher = patch('cleveragents.providers.llm.langchain_chat_provider.PlanGenerationGraph')
+    ...    invoke_patcher = patch('cleveragents.providers.llm.langchain_chat_provider.LangChainChatProvider._invoke_with_retry', side_effect=retry_error)
+    ...    chat_cls = chat_patcher.start()
+    ...    graph_cls = graph_patcher.start()
+    ...    invoke_patcher.start()
+    ...    chat_instance = MagicMock(name='ChatGoogleGenerativeAIInstance')
+    ...    chat_instance.get_num_tokens.return_value = 0
+    ...    chat_cls.return_value = chat_instance
+    ...    graph_instance = MagicMock(name='GraphInstance')
+    ...    graph_cls.return_value = graph_instance
+    ...    provider = GoogleChatProvider(api_key='sk-google-robot', model='gemini-2.0-flash')
+    ...    response = provider.generate_changes(project, plan, contexts)
+    ...    chat_patcher.stop()
+    ...    graph_patcher.stop()
+    ...    invoke_patcher.stop()
+    ...    assert response.error_message == 'deadline reached'
+    ...    assert response.changes == []
+    ...    print('google-provider-retry-error')
+     ${result}=    Run Process    ${PYTHON}    -c    ${script}
+     Log Process Failure    ${result}
+
+    Should Be Equal As Integers    ${result.rc}    0
+    Should Contain    ${result.stdout}    google-provider-retry-error
+
+*** Keywords ***
+Log Process Failure
+    [Arguments]    ${result}
+    Run Keyword If    ${result.rc} == 0    Return From Keyword
+    Log To Console    Process failed with rc=${result.rc}
+    Log To Console    STDOUT:${\n}${result.stdout}
+    Log To Console    STDERR:${\n}${result.stderr}

src/cleveragents/providers/llm/google_provider.py

@@ -0,0 +1,39 @@
+from __future__ import annotations
+
+from typing import Any
+
+from langchain_google_genai import ChatGoogleGenerativeAI
+
+from cleveragents.providers.llm.langchain_chat_provider import LangChainChatProvider
+
+
+class GoogleChatProvider(LangChainChatProvider):
+    """Concrete provider adapter for Google Gemini chat models."""
+
+    def __init__(
+        self,
+        *,
+        api_key: str,
+        model: str = "gemini-2.0-flash",
+        max_retries: int = 3,
+        **llm_kwargs: Any,
+    ) -> None:
+        if not api_key:
+            raise ValueError("Google API key is required")
+
+        def factory(resolved_model: str) -> ChatGoogleGenerativeAI:
+            kwargs: dict[str, Any] = {
+                "api_key": api_key,
+                "model": resolved_model,
+            }
+            if llm_kwargs:
+                kwargs.update(llm_kwargs)
+            return ChatGoogleGenerativeAI(**kwargs)
+
+        super().__init__(
+            name="google",
+            model_id=model,
+            llm_factory=factory,
+            max_retries=max_retries,
+            supports_streaming=True,
+        )

src/cleveragents/providers/registry.py

@@ -582,6 +582,27 @@ class ProviderRegistry:
             provider_info.capabilities if provider_info else ProviderCapabilities()
         )
 
+        if provider_type == ProviderType.GOOGLE:
+            from cleveragents.providers.llm.google_provider import GoogleChatProvider
+
+            key_attr = self.PROVIDER_KEY_ATTRS.get(provider_type)
+            api_key = getattr(self._settings, key_attr, None) if key_attr else None
+            if not api_key:
+                missing_env = (
+                    key_attr.upper() if key_attr else provider_type.value.upper()
+                )
+                raise ValueError(
+                    f"Provider {provider_type.value} is not configured. "
+                    f"Please set the {missing_env} environment variable."
+                )
+
+            return GoogleChatProvider(
+                api_key=api_key,
+                model=model_id
+                or self.DEFAULT_MODELS.get(provider_type, "gemini-2.0-flash"),
+                max_retries=max_retries,
+            )
+
         # Create factory function for the LLM
         def llm_factory(mid: str) -> BaseLanguageModel:
             return self._create_provider_llm(provider_type, mid)  # type: ignore[arg-type]