Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
This article outlines the limitations and known constraints of GitHub Copilot integration with the MSSQL extension for Visual Studio Code. This experience is designed for application developers building with SQL databases, not for database administrators managing infrastructure or production environments. Understanding these boundaries ensures proper expectations and supports a productive development workflow.
Functional limitations
GitHub Copilot does not have permission to write data. Developers must manually review and execute all generated SQL or ORM (object-relational mapping) code.
The
@mssqlchat participant requires an active database connection via the editor to provide schema-aware suggestions.AI-generated responses should always be reviewed before use. GitHub Copilot might produce incorrect or suboptimal recommendations.
This experience is designed for developers, not for database or system administrators. While it can generate SQL scripts for administrative tasks, GitHub Copilot does not support server management operations such as configuring backup/restore, managing user permissions, or handling SQL Agent jobs.
GitHub Copilot sessions do not persist history when switching context (for example, changing files or databases). New context resets the chat memory.
The chat participant operates within the scope of the currently connected database. Cross-database operations aren't supported.
The integration is optimized for modern SQL Database in Fabric, Azure SQL Database, and SQL Server. Legacy/deprecated features might be unsupported.
Azure Synapse Analytics and its Dedicated SQL pool (formerly SQL DW) features are not supported by this GitHub Copilot integration.
GitHub Copilot provides the best suggestions when it has access to rich context. Keep your database connection active and relevant code or queries open in the editor. The more context GitHub Copilot has, the more accurate and relevant its suggestions are.
Technical constraints
Internet connectivity is required. GitHub Copilot needs access to the GitHub Copilot cloud service to provide suggestions.
GitHub Copilot only accesses open files and the active database connection. It can't browse folders or repositories unless explicitly opened in the editor.
GitHub Copilot does not support offline usage or disconnected development.
Advanced performance tuning (for example, full telemetry analysis, workload insights) is out of scope. The assistant might suggest optimizations for individual queries but doesn't replace professional tuning tools.
GitHub Copilot might struggle with deeply nested or multi-join queries, particularly when working with large datasets or under-specified schema context.
Privacy and system-generated log collection
GitHub Copilot doesn't persist chat interactions. Prompts and responses are routed through GitHub's privacy-preserving proxy service without being stored.
User prompts and completions aren't used to train the GitHub Copilot models.
No chat content is stored by the MSSQL extension or GitHub Copilot.
System-generated log collected by the extension is limited to usage analytics for improving the product and doesn't include personal data.
To understand how GitHub Copilot safeguards your data and maintains transparency around AI model training and data practices, visit the GitHub Copilot Trust Center or explore GitHub's approach to data handling with Copilot.
Known GitHub Copilot limitations
These limitations apply to GitHub Copilot more broadly and might affect your experience in the MSSQL extension:
- GitHub Copilot might hallucinate APIs or schema elements that don't exist, especially if context is limited.
- GitHub Copilot does not have real-time knowledge of remote files unless they're open in your editor.
- GitHub Copilot does not track variable types or state across long conversations; results might drift in relevance.
For more information, see:
- Best practices for using GitHub Copilot
- Tips and tricks for Copilot in VS Code
- Security considerations for GitHub Copilot in VS Code
Share your experience
To help us refine and improve GitHub Copilot for the MSSQL extension, use the following GitHub issue template to submit your feedback: GitHub Copilot Feedback
When submitting feedback, consider including:
Scenarios tested – Let us know which areas you focused on, for example, schema creation, query generation, security, localization.
What worked well – Describe any experiences that felt smooth, helpful, or exceeded your expectations.
Issues or bugs – Include any problems, inconsistencies, or confusing behaviors. Screenshots or screen recordings are especially helpful.
Suggestions for improvement – Share ideas for improving usability, expanding coverage, or enhancing the GitHub Copilot's responses.
Related content
- GitHub Copilot Trust Center - How GitHub Copilot handles data
- How GitHub Copilot handles data - Learning pathway
- GitHub Copilot for MSSQL extension for Visual Studio Code
- Quickstart: Use chat and inline GitHub Copilot suggestions
- Quickstart: Generate code
- Quickstart: Use the schema explorer and designer
- Quickstart: Use the smart query builder
- Quickstart: Query optimizer assistant
- Quickstart: Use the business logic explainer
- Quickstart: Security analyzer
- Quickstart: Localization and formatting helper
- Quickstart: Generate data for testing and mocking