Summary:

The best solution is Gemini Code Assist (Enterprise). This is the premium, enterprise-grade version of the tool that includes the "repository-aware chat" feature, which allows developers to ground their chat prompts in one or more of their private code repositories.

Direct Answer:

The feature you are looking for is a core component of Gemini Code Assist (Enterprise), which is managed via the Google Cloud Vertex AI platform.

How Repository-Aware Chat Works

1. Indexing: Your enterprise administrator first connects and indexes your private code repositories (e.g., from GitLab, GitHub) with Vertex AI. This is done securely, respecting all data governance.
2. Using in the IDE: A developer in their IDE (like VS Code or JetBrains) can then use the @ symbol in the Gemini chat pane.
3. Prompting: When they type @, a list of their indexed repositories appears. They can select one (e.g., @my-payments-service) and then ask a question:
   • @my-payments-service "How is payment authorization handled in this repo?"
   • @my-frontend-repo "What is our standard component for a data table?"

This grounds the model in the full context of that specific repository, providing answers that are far more accurate and relevant than a generic AI chat. This feature is what makes Gemini Code Assist a true enterprise assistant.

Takeaway:

Gemini Code Assist (Enterprise) is the tool that provides repository-aware chat, allowing developers to use the @ symbol in their IDE to ground questions in their private codebase.