Add support for Anthropic/Claude and other LLM providers #74

dcpagotto · 2026-01-31T04:17:52Z

Hey, been using ROSA for some robotics projects and ran into the limitation of it being locked to OpenAI only. Saw that other people had the same problem (#56, #65) so I decided to refactor the agent layer to support multiple providers.

What I did

The main change is swapping out the OpenAI-specific agent pipeline (format_to_openai_tool_messages + OpenAIToolsAgentOutputParser) for langchain's create_tool_calling_agent, which works with any model that has tool calling support. This means you can now use Claude, Gemini, local models through Ollama, etc.

I also:

Changed ChatModel type to BaseChatModel so it accepts any langchain chat model
Added a guard on the token usage callback so it doesnt crash when using non-OpenAI models
Updated the turtle_agent example so you can switch providers with an env var
Added langchain-anthropic as dependency and updated .env with config for all providers

How to use it

# OpenAI (same as before, nothing changes)
LLM_PROVIDER=openai OPENAI_API_KEY=sk-... python turtle_agent.py

# Anthropic
LLM_PROVIDER=anthropic ANTHROPIC_API_KEY=sk-ant-... python turtle_agent.py

# Ollama
LLM_PROVIDER=ollama OLLAMA_MODEL=llama3 python turtle_agent.py

Backwards compatible

Nothing breaks. If you're already using ChatOpenAI or AzureChatOpenAI everything stays the same.

Relates to #56, #65

Refactored the agent construction to be provider-agnostic by replacing the OpenAI-specific tool message formatter and output parser with langchain's create_tool_calling_agent, which works with any chat model that implements tool calling. Changes: - Replace format_to_openai_tool_messages + OpenAIToolsAgentOutputParser with create_tool_calling_agent (provider-agnostic) - Widen ChatModel type alias to BaseChatModel so any langchain-compatible LLM can be passed in without type errors - Add langchain-anthropic as an optional dependency - Guard token usage tracking behind an isinstance check so it only runs the OpenAI callback for OpenAI models (avoids runtime errors with other providers) - Update turtle_agent example to support provider selection via LLM_PROVIDER env var (openai | anthropic | ollama) - Update .env with configuration for all three providers Backwards compatible: existing code using ChatOpenAI or AzureChatOpenAI will work exactly the same. No breaking changes to the public API. Relates to nasa-jpl#56

Copilot

Pull request overview

This pull request adds support for multiple LLM providers (Anthropic/Claude, Ollama) to ROSA, which was previously locked to OpenAI only. The main architectural change replaces OpenAI-specific agent components (format_to_openai_tool_messages and OpenAIToolsAgentOutputParser) with langchain's provider-agnostic create_tool_calling_agent function.

Changes:

Refactored the agent layer in ROSA to use provider-agnostic create_tool_calling_agent instead of OpenAI-specific components
Added provider selection logic to turtle_agent/scripts/llm.py with support for OpenAI, Anthropic, and Ollama
Updated the ChatModel type from a Union of specific models to BaseChatModel for broader compatibility
Added conditional token usage tracking that only works for OpenAI-based models
Updated dependencies and environment configuration to support multiple providers

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 10 comments.

File	Description
src/rosa/rosa.py	Refactored agent creation to use `create_tool_calling_agent`, updated type annotations to accept any `BaseChatModel`, and added conditional token usage tracking for OpenAI models
src/turtle_agent/scripts/llm.py	Added provider switching logic based on `LLM_PROVIDER` environment variable with support for OpenAI, Anthropic, and Ollama
pyproject.toml	Added `langchain-anthropic~=0.3.12` as a dependency
.env	Added configuration sections for all three providers with model names and API keys

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-01-31T22:37:28Z

src/turtle_agent/scripts/llm.py

 def get_llm(streaming: bool = False):
-    """A helper function to get the LLM instance."""
+    """A helper function to get the LLM instance.
+
+    Supports OpenAI (default), Anthropic and Ollama models.
+    Set the LLM_PROVIDER env variable to switch between providers:
+      - "openai" (default): uses OPENAI_API_KEY
+      - "anthropic": uses ANTHROPIC_API_KEY
+      - "ollama": uses local Ollama instance
+    """
    dotenv.load_dotenv(dotenv.find_dotenv())

-    llm = ChatOpenAI(
-        api_key=get_env_variable("OPENAI_API_KEY"),
-        model="gpt-5.1",
-        streaming=streaming,
-    )
+    provider = os.getenv("LLM_PROVIDER", "openai").lower()
+
+    if provider == "anthropic":
+        try:
+            from langchain_anthropic import ChatAnthropic
+        except ImportError:
+            raise ImportError(
+                "langchain-anthropic is required for Anthropic support. "
+                "Install it with: pip install langchain-anthropic"
+            )
+        llm = ChatAnthropic(
+            api_key=get_env_variable("ANTHROPIC_API_KEY"),
+            model=os.getenv("ANTHROPIC_MODEL", "claude-sonnet-4-20250514"),
+            streaming=streaming,
+        )
+    elif provider == "ollama":
+        from langchain_ollama import ChatOllama
+        llm = ChatOllama(
+            model=os.getenv("OLLAMA_MODEL", "llama3"),
+            base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),
+        )
+    else:
+        llm = ChatOpenAI(
+            api_key=get_env_variable("OPENAI_API_KEY"),
+            model=os.getenv("OPENAI_MODEL", "gpt-4o"),
+            streaming=streaming,
+        )

    return llm


The new multi-provider functionality lacks test coverage. The codebase has comprehensive test coverage for tools and other components, but there are no tests validating that the different LLM providers (Anthropic, Ollama) can be instantiated correctly, or that the provider switching logic works as expected. Consider adding tests that verify the get_llm function with different LLM_PROVIDER values and mock the provider-specific imports.

src/rosa/rosa.py

Copilot · 2026-01-31T22:37:29Z

pyproject.toml

    "langchain-core~=0.3.52",
    "langchain-openai~=0.3.14",
    "langchain-ollama~=0.3.2",
+    "langchain-anthropic~=0.3.12",


The langchain-anthropic package is listed as a required dependency in pyproject.toml (line 33), but the code in llm.py treats it as optional with a try-except ImportError block. This creates an inconsistency: either make langchain-anthropic an optional dependency in pyproject.toml, or remove the try-except block since the package will always be installed. The current approach will always install langchain-anthropic even if users only want to use OpenAI or Ollama, which increases installation size and dependencies unnecessarily.

Copilot · 2026-01-31T22:37:30Z

src/rosa/rosa.py

+from langchain_core.language_models import BaseChatModel
 from langchain_core.messages import AIMessage, HumanMessage
 from langchain_core.prompts import ChatPromptTemplate
 from langchain_ollama import ChatOllama


The ChatOllama import is no longer used in this file after refactoring to use BaseChatModel. The import can be safely removed as ChatOllama instances will be passed in as BaseChatModel types. Only ChatOpenAI and AzureChatOpenAI are needed for the token callback isinstance check.

Suggested change

from langchain_ollama import ChatOllama

Copilot · 2026-01-31T22:37:30Z

src/rosa/rosa.py

+    try:
+        from langchain_anthropic import ChatAnthropic
+    except ImportError:
+        pass


The TYPE_CHECKING import block for ChatAnthropic will fail silently if the import fails, but this won't provide useful type hints since the import is wrapped in a try-except. Consider removing the try-except within the TYPE_CHECKING block, as import errors during type checking should be reported to developers so they know to install the appropriate dependencies for type checking. The runtime import protection is already handled in the turtle_agent/scripts/llm.py file.

Suggested change

try:

from langchain_anthropic import ChatAnthropic

except ImportError:

pass

from langchain_anthropic import ChatAnthropic

Copilot · 2026-01-31T22:37:30Z

src/rosa/rosa.py

        ros_version (Literal[1, 2]): The version of ROS that the agent will interact with.
-        llm (Union[AzureChatOpenAI, ChatOpenAI, ChatOllama]): The language model to use for generating responses.
+        llm (BaseChatModel): Any langchain chat model that supports tool calling. Tested with
+            ChatOpenAI, AzureChatOpenAI, ChatOllama, and ChatAnthropic.


The llm parameter documentation should mention that token usage tracking via show_token_usage is only available for ChatOpenAI and AzureChatOpenAI models. When using other providers like ChatAnthropic or ChatOllama, token usage will not be displayed even if show_token_usage is enabled.

Suggested change

ChatOpenAI, AzureChatOpenAI, ChatOllama, and ChatAnthropic.

ChatOpenAI, AzureChatOpenAI, ChatOllama, and ChatAnthropic. Note that token usage

tracking via `show_token_usage` is only supported for ChatOpenAI and AzureChatOpenAI;

other providers (e.g., ChatAnthropic, ChatOllama) will not display token usage even

when `show_token_usage` is enabled.

Copilot · 2026-01-31T22:37:30Z

src/turtle_agent/scripts/llm.py

+            streaming=streaming,
+        )
+    elif provider == "ollama":
+        from langchain_ollama import ChatOllama


The langchain_ollama import is not wrapped in a try-except block like the langchain_anthropic import, creating inconsistent error handling between providers. Since both are listed as required dependencies in pyproject.toml, the ImportError handling for Anthropic is unnecessary. However, if the intention is to make these providers optional, then both should have try-except blocks with helpful error messages, and both should be made optional dependencies in pyproject.toml.

Suggested change

from langchain_ollama import ChatOllama

try:

from langchain_ollama import ChatOllama

except ImportError:

raise ImportError(

"langchain-ollama is required for Ollama support. "

"Install it with: pip install langchain-ollama"

)

Copilot · 2026-01-31T22:37:31Z

src/turtle_agent/scripts/llm.py

+        from langchain_ollama import ChatOllama
+        llm = ChatOllama(
+            model=os.getenv("OLLAMA_MODEL", "llama3"),
+            base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),


The streaming parameter is not passed to ChatOllama, unlike ChatOpenAI and ChatAnthropic. This inconsistency means that streaming won't work properly when using Ollama. Add the streaming parameter to maintain consistent behavior across all providers.

Suggested change

base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),

base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),

streaming=streaming,

Copilot · 2026-01-31T22:37:31Z

src/rosa/rosa.py

-from langchain.agents.output_parsers.openai_tools import OpenAIToolsAgentOutputParser
+import logging
+from contextlib import contextmanager
+from typing import TYPE_CHECKING, Any, AsyncIterable, Dict, Literal, Optional, Union


Import of 'Union' is not used.

Suggested change

from typing import TYPE_CHECKING, Any, AsyncIterable, Dict, Literal, Optional, Union

from typing import TYPE_CHECKING, Any, AsyncIterable, Dict, Literal, Optional

RobRoyce

Thanks for the PR! Overall this is a great addition and solves open requests for Anthropic model support. There are a few minor tweaks needed to ensure there is no confusion about which model providers are supported (especially when we move from hard coding to using env vars).

RobRoyce · 2026-01-31T22:36:44Z

.env

+
+# Anthropic Configuration
+ANTHROPIC_API_KEY=
+ANTHROPIC_MODEL=claude-sonnet-4-20250514


Likely to break as model snapshots are retired, prefer leaving blank (placeholders are fine).

Suggested change

ANTHROPIC_MODEL=claude-sonnet-4-20250514

ANTHROPIC_MODEL=claude-sonnet-4-5

RobRoyce · 2026-01-31T22:42:03Z

src/rosa/rosa.py

+logger = logging.getLogger(__name__)
+
+# Runtime-safe type alias: accepts any BaseChatModel, covering OpenAI, Azure,
+# Anthropic, Ollama and any future langchain provider that implements tool calling.
+ChatModel = BaseChatModel


The strict Union was removed in favor of a generic BaseChatModel, but get_llm()in llm.py still effectively hard-codes a strict union.

I am in favor of allowing any tool calling LLM, but that means we need the ability to construct a generic chat model without enumerating all possibilities.

Recommend either reverting back to Union with ChatAnthropic included and remove "any tool calling model" language, or find a good solution for constructing generic tool calling Chat models. Either fix is fine with me 👍

RobRoyce · 2026-01-31T22:50:53Z

src/rosa/rosa.py

+    @contextmanager
+    def _token_callback(self):
+        """Context manager for token usage tracking.
+
+        Uses the OpenAI callback when the LLM is an OpenAI-based model,
+        otherwise yields None so the rest of the flow is unaffected.
+        """
+        if isinstance(self.__llm, (ChatOpenAI, AzureChatOpenAI)):
+            with get_openai_callback() as cb:
+                yield cb
+        else:
+            yield None
+
    def _print_usage(self, cb):
        """Print the token usage if show_token_usage is enabled."""
-        if cb and self.__show_token_usage:
-            print(f"[bold]Prompt Tokens:[/bold] {cb.prompt_tokens}")
-            print(f"[bold]Completion Tokens:[/bold] {cb.completion_tokens}")
-            print(f"[bold]Total Cost (USD):[/bold] ${cb.total_cost}")
+        if cb is None or not self.__show_token_usage:
+            return
+        print(f"[bold]Prompt Tokens:[/bold] {cb.prompt_tokens}")
+        print(f"[bold]Completion Tokens:[/bold] {cb.completion_tokens}")
+        print(f"[bold]Total Cost (USD):[/bold] ${cb.total_cost}")


Silent inaction is undesirable when the user explicitly enables __show_token_usage (default is False). Prefer a small warning message and/or failing early (i.e. during __init__).

Question: does LangChain provide the ability to view tokens for arbitrary chat models? Since TikToken is already a dependency, perhaps we can use it to find token count for prompt + response?

RobRoyce · 2026-01-31T22:53:51Z

.env

 OPENAI_API_VERSION=
-OPENAI_API_TYPE=
+OPENAI_API_TYPE=
+OPENAI_MODEL=gpt-4o


Might as well go with the latest if we're adding defaults :)

Suggested change

OPENAI_MODEL=gpt-4o

OPENAI_MODEL=gpt-5.2

RobRoyce · 2026-01-31T23:00:25Z

src/turtle_agent/scripts/llm.py

I like the LLM_PROVIDER env var idea, but I don't think it should default to OpenAI if it doesn't match anthropic or ollama. If we are going to hard-code these names, prefer also hard-coding openai in the if/else and failing if there is no match to avoid confusion (error message should include the supported options).

RobRoyce · 2026-01-31T23:02:15Z

src/turtle_agent/scripts/llm.py

-        model="gpt-5.1",
-        streaming=streaming,
-    )
+    provider = os.getenv("LLM_PROVIDER", "openai").lower()


Please use the get_env_variable function and remove defaults (likely to cause confusion if this doesn't fail early with explicit error messages).

…ic dep, token usage warnings

Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 8 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

src/turtle_agent/scripts/llm.py

Copilot · 2026-02-01T17:02:56Z

src/turtle_agent/scripts/llm.py

+    if provider == "openai":
+        llm = ChatOpenAI(
+            api_key=get_env_variable("OPENAI_API_KEY"),
+            model=get_env_variable("OPENAI_MODEL"),


The PR description claims this change is "backwards compatible" and that "Nothing breaks," but the change from a hardcoded model to requiring the OPENAI_MODEL environment variable is a breaking change for turtle_agent users. Existing users who upgrade will encounter errors if they don't update their .env file. This should be documented as a breaking change in the PR description or a migration guide should be provided.

Copilot · 2026-02-01T17:02:56Z

src/rosa/rosa.py

    def _get_agent(self):
-        """Create and return an agent for processing user inputs and generating responses."""
-        agent = (
-            {
-                "input": lambda x: x["input"],
-                "agent_scratchpad": lambda x: format_to_openai_tool_messages(
-                    x["intermediate_steps"]
-                ),
-                "chat_history": lambda x: x["chat_history"],
-            }
-            | self.__prompts
-            | self.__llm_with_tools
-            | OpenAIToolsAgentOutputParser()
+        """Create and return an agent for processing user inputs and generating responses.
+
+        Uses create_tool_calling_agent which is provider-agnostic and works with
+        any LLM that supports tool calling (OpenAI, Anthropic, Ollama, etc).
+        """
+        agent = create_tool_calling_agent(
+            llm=self.__llm,
+            tools=self.__tools.get_tools(),
+            prompt=self.__prompts,
        )
        return agent


The refactoring from OpenAI-specific agent pipeline to create_tool_calling_agent is a significant architectural change, but there are no tests to verify that it works correctly with different providers. Consider adding integration tests that validate the agent works with mock ChatOpenAI, ChatAnthropic, and ChatOllama instances to ensure the provider-agnostic implementation behaves correctly across different LLM backends.

Copilot · 2026-02-01T17:02:56Z

src/rosa/rosa.py

+if TYPE_CHECKING:
+    from langchain_anthropic import ChatAnthropic
+


The ChatAnthropic import under TYPE_CHECKING is not used in any type annotations, only in the docstring. Since the code now uses BaseChatModel for all type hints, this import can be removed to reduce dependencies and avoid potential confusion.

Suggested change

if TYPE_CHECKING:

from langchain_anthropic import ChatAnthropic

Copilot · 2026-02-01T17:02:56Z

pyproject.toml

+[project.optional-dependencies]
+anthropic = ["langchain-anthropic~=0.3.12"]


Consider adding langchain-ollama to the optional-dependencies section similar to langchain-anthropic, since it's no longer imported directly in the ROSA core code and is only used in the turtle_agent example with lazy import. This would reduce the installation footprint for users who don't need Ollama support. For example: ollama = ["langchain-ollama~=0.3.2"]

Copilot · 2026-02-01T17:02:57Z

src/rosa/rosa.py

+            if self.__show_token_usage:
+                logger.warning("Token usage tracking is only supported for OpenAI and Azure models.")


The warning message is only logged once when the context manager is entered, but if a user repeatedly calls invoke() with show_token_usage=True on a non-OpenAI model, they will see this warning every time. Consider adding a flag to log this warning only once, or document this limitation in the constructor's docstring more prominently to set user expectations upfront.

Copilot · 2026-02-01T17:02:57Z

src/rosa/rosa.py

+        if isinstance(self.__llm, (ChatOpenAI, AzureChatOpenAI)):
+            with get_openai_callback() as cb:
+                yield cb
+        else:
+            if self.__show_token_usage:
+                logger.warning("Token usage tracking is only supported for OpenAI and Azure models.")


The _token_callback method uses isinstance() to check if the LLM is OpenAI-based, but this check happens at runtime on every invocation. Since the LLM type doesn't change after initialization, consider checking this once in init and storing a boolean flag (e.g., self.__supports_token_tracking) to avoid repeated isinstance calls. This would be more efficient and clearer.

Suggested change

if isinstance(self.__llm, (ChatOpenAI, AzureChatOpenAI)):

with get_openai_callback() as cb:

yield cb

else:

if self.__show_token_usage:

logger.warning("Token usage tracking is only supported for OpenAI and Azure models.")

# Lazily determine whether the current LLM supports OpenAI-style token tracking.

if not hasattr(self, "_ROSA__supports_token_tracking"):

self.__supports_token_tracking = isinstance(

self.__llm, (ChatOpenAI, AzureChatOpenAI)

)

if self.__supports_token_tracking:

with get_openai_callback() as cb:

yield cb

else:

if self.__show_token_usage:

logger.warning(

"Token usage tracking is only supported for OpenAI and Azure models."

)

dcpagotto · 2026-02-01T18:43:31Z

hey @RobRoyce, thanks for the detailed review! pushed the fixes:

Union vs BaseChatModel: went with a TYPE_CHECKING approach -- type checkers see the explicit Union[ChatOpenAI, AzureChatOpenAI, ChatAnthropic, ChatOllama], runtime keeps BaseChatModel since anthropic/ollama are optional deps. updated the docstring to list tested providers and droped the "any tool calling model" wording.

Token usage: moved the warning to __init__ so it fails early when show_token_usage=True with a non-OpenAI model, includes the model class name so its clear whats going on. removed the per-call warning from _token_callback since the init check covers it. re: your question about langchain token counting -- get_openai_callback is OpenAI-only, other providers would need hooks into tiktoken or their own usage API. could be a good follow-up PR.

btw, been using ROSA with a hexapod robot i have (JetHEXA -- Jetson Nano B01, ROS Melodic, 18 DOF with IK, LIDAR, depth camera, IMU). the multi-provider support is really usefull here -- been testing with both OpenAI and Claude to compare how they handle ROS service calls and tool chaining.

the Ollama support is actually the one im most excited about tho. the whole point for me is running inference directly on the Jetson so the robot can operate fully offline -- no cloud dependency, no latency, just edge AI. think autonomous decision-making beyond what traditional nav stacks give you (SLAM, AMCL, move_base). similar to how JPL does it with AutoNav on the Mars rovers or AEGIS for autonomous target selection -- the robot needs to reason on its own when theres no link back home.

im also planing a custom ROSA tool called learn where the agent would iteratively command joint angles across the 18 servos, read back IMU stability + servo position feedback from ROS topics, and progressively build optimized gait patterns -- starting with stair climbing. basically using the LLM as a high-level controller for locomotion learning. will def share once i have something working

let me know if the fixes look good

dcpagotto force-pushed the feat/anthropic-support branch from e530585 to 56adc8b Compare January 31, 2026 04:26

dcpagotto changed the title ~~feat: add Anthropic/Claude support and make agent provider-agnostic~~ feat: add multi-provider LLM support (Anthropic, Google, etc.) Jan 31, 2026

dcpagotto changed the title ~~feat: add multi-provider LLM support (Anthropic, Google, etc.)~~ Add support for Anthropic/Claude and other LLM providers Jan 31, 2026

This was referenced Jan 31, 2026

[New Feature]: will Anthropic api be supported in the future? Models like sonnet 3.5/3.7 etc. #56

Closed

[New Feature]: Is there a compatible version to work with ollama ? #65

Closed

RobRoyce requested a review from Copilot January 31, 2026 22:31

Copilot started reviewing on behalf of RobRoyce January 31, 2026 22:31 View session

Copilot AI reviewed Jan 31, 2026

View reviewed changes

RobRoyce requested changes Jan 31, 2026

View reviewed changes

address review feedback: strict provider validation, optional anthrop…

788083b

…ic dep, token usage warnings

RobRoyce requested a review from Copilot February 1, 2026 16:57

Copilot started reviewing on behalf of RobRoyce February 1, 2026 16:57 View session

Copilot AI reviewed Feb 1, 2026

View reviewed changes

address review: explicit provider union, early token usage warning

7723d5d

-            ChatOpenAI, AzureChatOpenAI, ChatOllama, and ChatAnthropic.
+            ChatOpenAI, AzureChatOpenAI, ChatOllama, and ChatAnthropic. Note that token usage
+            tracking via `show_token_usage` is only supported for ChatOpenAI and AzureChatOpenAI;
+            other providers (e.g., ChatAnthropic, ChatOllama) will not display token usage even
+            when `show_token_usage` is enabled.

	base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),
	base_url=os.getenv("OLLAMA_BASE_URL", "http://localhost:11434"),
	streaming=streaming,

	from typing import TYPE_CHECKING, Any, AsyncIterable, Dict, Literal, Optional, Union
	from typing import TYPE_CHECKING, Any, AsyncIterable, Dict, Literal, Optional

	ANTHROPIC_MODEL=claude-sonnet-4-20250514
	ANTHROPIC_MODEL=claude-sonnet-4-5

		if TYPE_CHECKING:
		from langchain_anthropic import ChatAnthropic

		[project.optional-dependencies]
		anthropic = ["langchain-anthropic~=0.3.12"]

		if self.__show_token_usage:
		logger.warning("Token usage tracking is only supported for OpenAI and Azure models.")

Add support for Anthropic/Claude and other LLM providers #74

Are you sure you want to change the base?

Add support for Anthropic/Claude and other LLM providers #74

Uh oh!

Conversation

dcpagotto commented Jan 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What I did

How to use it

Backwards compatible

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce left a comment

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

RobRoyce Jan 31, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 1, 2026

dcpagotto commented Jan 31, 2026 •

edited

Loading