Search

FAQ

General Questions

What is CodeComet’s App Copilot?

App Copilot by CodeComet is an AI-driven tool that helps engineering teams quickly diagnose and resolve issues in production APIs and backend systems. By integrating with telemetry and version control systems, App Copilot uses Generative AI to analyze errors and performance anomalies, providing code-level fix suggestions to reduce downtime and improve system reliability.

How does App Copilot improve my team’s workflow?

App Copilot reduces the manual effort involved in debugging production issues by automatically detecting errors and performance bottlenecks, analyzing the context, and suggesting specific code fixes. This allows engineers to focus on high-impact tasks like new features and architectural improvements, while App Copilot handles the heavy lifting in issue resolution.

Which telemetry and version control systems are currently supported?

Currently, App Copilot integrates with Datadog (APM and Logs) for telemetry data and GitHub for version control. We’re eager to hear from customers about the tools they use, so please let us know if you’d like support for other systems like New Relic, Sentry, GitLab, or Anthropic.

Installation and Setup

How do I set up App Copilot with Datadog and GitHub?

To begin using App Copilot, connect your Datadog and GitHub accounts through the CodeComet dashboard. This will enable App Copilot to parse telemetry data from Datadog and access your codebase on GitHub. Note that some manual setup may be required, especially to configure telemetry parsing and codebase interfacing before the error and performance analyses start streaming in.

Will my whole team need to configure access in GitHub?

No, App Copilot utilizes existing GitHub permissions for seamless onboarding. Once an admin sets up a project in App Copilot, all team members with access to the associated GitHub repository automatically gain access to the project within CodeComet.

Configuration and Use

What types of issues does App Copilot detect?

App Copilot detects errors such as 4xx and 5xx server responses, as well as performance anomalies related to resource consumption and response times. It captures stack traces, identifies relevant code paths, and forwards this context to an LLM to generate actionable code fixes or optimization suggestions.

Can I choose the LLM provider App Copilot uses?

Currently, App Copilot uses OpenAI for LLM analysis. However, we aim to support a range of LLM providers to fit customer needs. If you’re interested in using a different LLM provider, we’d love to hear from you.

How does App Copilot handle repeated issues?

App Copilot intelligently groups similar issues, so recurring problems are only analyzed once. This provides your team with a clear count of how often an issue occurs and prevents redundant analysis, streamlining the troubleshooting process.

Collaboration and Notifications

How will App Copilot notify my team about issues?

App Copilot integrates with collaboration platforms like Slack, Discord, and Google Chat, so your team receives issue alerts directly in their preferred workspace. In addition, you can view suggested fixes and insights through the CodeComet web app.

Can I create branches or pull requests from App Copilot’s suggested fixes?

Yes, App Copilot allows you to create a new branch or PR in GitHub based on suggested code changes, making it easy to apply, review, or modify fixes directly from CodeComet.

Troubleshooting

What if I encounter issues with setup or configuration?

If you encounter any setup issues, please ensure your Datadog and GitHub accounts are properly linked in CodeComet and that telemetry parsing is correctly configured. If problems persist, reach out to our support team for assistance.

Advanced Use

Is App Copilot suitable for use with additional programming languages or frameworks?

At the moment, App Copilot supports Python and Django. We are actively exploring additional languages and frameworks and would love to hear from you if your stack includes other languages and frameworks like TypeScript, JavaScript, Ruby, Flask, or FastAPI.

What are CodeComet’s plans for expanding App Copilot’s capabilities?

We’re committed to continuously improving App Copilot. Planned enhancements include support for additional telemetry and version control systems, more language and framework options, and greater flexibility in choosing LLM providers. Our goal is to make App Copilot an invaluable tool for teams working across diverse stacks.

How can I request a feature for App Copilot?

We value your feedback. Please contact us or email support@codecomet.io to request specific features or integrations that would help App Copilot better serve your team’s needs.