Examples of AI Code Reviews

See Bito’s AI Code Review Agent in action

Real-world examples of how AI improves your code reviews

Supported languages: C, C++, Java, JavaScript, PHP, Python, & more

Supported languages:
C, C++, Java, JavaScript, PHP, Python, & more

Choose a language to see examples (More examples coming soon)

Choose a language to see examples
(More examples coming soon)

Example 1:

Agent fixes issues in PR, streamlining the merge for Devika

Devika is an open-source alternative to Devin AI, an AI software engineering agent powered by large language models (LLM).
Bito’s AI Code Review Agent identified issues in a pull request that aimed to enable the configuration of the OpenAI API Base URL and to integrate DuckDuckGo as an alternative search engine alongside Bing.

Suggestion 1:

The Agent suggested a fix for a potential issue where the code might not handle empty or missing values for the base_url parameter during OpenAI client initialization.

				
					if not base_url:
    base_url = 'https://api.openai.com/v1'
self.client = OAI(
    api_key=api_key,
    base_url=base_url
)

Suggestion 2:

See how the Agent spotted type inconsistencies in the DuckDuckGoSearch class’s Search method, which could trigger runtime errors when consuming the method’s output.

				
					try:
    self.query_result = DDGS().text(query, max_results=5)
    return self.query_result
except Exception as err:
    # Log the error
    print(f'Error occurred: {err}')
    return []

Example 1:

AI identifies memory leak issue in Java code

Apache Dubbo is a high-performance, Java-based open-source RPC and microservice framework. Bito’s AI Code Review Agent, capable of understanding Java code, identified issues in a pull request focusing on fixing a memory leak and adding new features. The Agent recommends adding unit tests for new code paths and ensuring thread safety to enhance code reliability.

Suggestion 1:

The Agent identified a potential resource leak in the Http2ServerChannelObserver class, where the closed flag is set without ensuring resource cleanup, which may cause performance issues.

				
					if (closed) {
    releaseAssociatedResources(); // Implement this method to clean up resources
}

Suggestion 2:

The Agent suggested validating the state of getStreamingDecoder() before invoking close() to avoid potential NullPointerException , which could cause runtime errors.

				
					if (getStreamingDecoder() != null) {
    getStreamingDecoder().close();
}

Example 2:

Identify scalability or performance issues early and often

Netty is an asynchronous event-driven network application framework for rapid development of maintainable high performance protocol servers and clients.  Bito’s AI Code Review Agent identified issues in a pull request that refactors AbstractScheduledEventExecutor to use a pluggable TaskScheduler interface for flexibility. The Agent suggested adding a TaskSchedulerFactory for dynamic creation and commended the new tests.

Suggestion 1:

See how the Agent addressed a scalability issue by suggesting an object pool for RunnableScheduledFuture instances instead of creating new ones, significantly improving performance by reducing object creation and garbage collection.

				
					private final ObjectPool<RunnableScheduledFuture<?>> futurePool;
public AbstractTaskScheduler(ObjectPool<RunnableScheduledFuture<?>> futurePool) {
    this.futurePool = futurePool;
}

@Override
public Future<Void> schedule(Runnable command, long delay, TimeUnit unit) {
    RunnableScheduledFuture<Void> task = futurePool.get();
    task.reset(command, delay, unit);
    return schedule(task);
}

Suggestion 2:

The Agent identified a performance issue in DefaultTaskScheduler , suggesting a custom approach for comparing RunnableScheduledFutureNode instances to enhance efficiency in managing scheduled tasks, particularly beneficial for large-scale operations.

				
					private static final Comparator<RunnableScheduledFuture<?>> TASK_SCHEDULER_COMPARATOR = new Comparator<RunnableScheduledFuture<?>>() {
    @Override
    public int compare(RunnableScheduledFuture<?> o1, RunnableScheduledFuture<?> o2) {
        return Long.compare(o1.getDelay(TimeUnit.NANOSECONDS), o2.getDelay(TimeUnit.NANOSECONDS));
    }
};

private PriorityQueue<RunnableScheduledFuture<?>> scheduledTaskQueue = new PriorityQueue<>(TASK_SCHEDULER_COMPARATOR);

Example 1:

Agent fixes issues in PR, streamlining the merge for Devika

Suggestion 1:

The Agent suggested a fix for a potential issue where the code might not handle empty or missing values for the base_url parameter during OpenAI client initialization.

				
					if not base_url:
    base_url = 'https://api.openai.com/v1'
self.client = OAI(
    api_key=api_key,
    base_url=base_url
)

Suggestion 2:

See how the Agent spotted type inconsistencies in the DuckDuckGoSearch class’s Search method, which could trigger runtime errors when consuming the method’s output.

				
					try:
    self.query_result = DDGS().text(query, max_results=5)
    return self.query_result
except Exception as err:
    # Log the error
    print(f'Error occurred: {err}')
    return []

Example 1:

AI identifies memory leak issue in Java code

Suggestion 1:

The Agent identified a potential resource leak in the Http2ServerChannelObserver class, where the closed flag is set without ensuring resource cleanup, which may cause performance issues.

				
					if (closed) {
    releaseAssociatedResources(); // Implement this method to clean up resources
}

Suggestion 2:

The Agent suggested validating the state of getStreamingDecoder() before invoking close() to avoid potential NullPointerException , which could cause runtime errors.

				
					if (getStreamingDecoder() != null) {
    getStreamingDecoder().close();
}

Example 2:

Identify scalability or performance issues early and often

Suggestion 1:

				
					private final ObjectPool<RunnableScheduledFuture<?>> futurePool;
public AbstractTaskScheduler(ObjectPool<RunnableScheduledFuture<?>> futurePool) {
    this.futurePool = futurePool;
}

@Override
public Future<Void> schedule(Runnable command, long delay, TimeUnit unit) {
    RunnableScheduledFuture<Void> task = futurePool.get();
    task.reset(command, delay, unit);
    return schedule(task);
}

Suggestion 2:

				
					private static final Comparator<RunnableScheduledFuture<?>> TASK_SCHEDULER_COMPARATOR = new Comparator<RunnableScheduledFuture<?>>() {
    @Override
    public int compare(RunnableScheduledFuture<?> o1, RunnableScheduledFuture<?> o2) {
        return Long.compare(o1.getDelay(TimeUnit.NANOSECONDS), o2.getDelay(TimeUnit.NANOSECONDS));
    }
};

private PriorityQueue<RunnableScheduledFuture<?>> scheduledTaskQueue = new PriorityQueue<>(TASK_SCHEDULER_COMPARATOR);

See Bito’s AI Code Review Agent in action

Agent fixes issues in PR, streamlining the merge for Devika

AI identifies memory leak issue in Java code

Identify scalability or performance issues early and often

Agent fixes issues in PR, streamlining the merge for Devika

AI identifies memory leak issue in Java code

Identify scalability or performance issues early and often

Ready for better code reviews?

Cut review time by 50%