What happened
GitHub has updated its Copilot plugin for JetBrains, introducing new administrative controls for model selection, streamlined diagnostic workflows, and manual management of Model Context Protocol (MCP) servers. The update also mandates a minimum IDE version of 2025.2, ending support for the 2025.1 release.
The update introduces a managed settings for enterprise administrators, allowing them to define a default Copilot agent model for new conversations. While this sets a standard starting point, the plugin retains the existing model picker, allowing individual developers to override the default if necessary.
Diagnostic workflows have been enhanced with a new 'Fix' action within intention menus. When a diagnostic issue is identified, developers can trigger this action to open an inline chat window where Copilot attempts to repair the code. The system automatically selects 'agent mode' if available, or defaults to 'ask mode' to facilitate the repair.
A new configuration setting allows users to disable the automatic startup of MCP servers for both Copilot and Claude. This provides developers with manual control over when these tools initialize, which can help manage IDE resource consumption and prevent unwanted background activity.
The update includes broad reliability improvements targeting language server startup, model switching, and worktree workflows. Additionally, GitHub has officially ended support for JetBrains IDE 2025.1, requiring users to upgrade to version 2025.2 or later to continue using the plugin.
Why it matters
These changes provide enterprise administrators with greater governance over AI model usage, ensuring that organizations can standardize the default AI experience for developers. By integrating diagnostic 'Fix' actions directly into the IDE, the update reduces the friction between identifying code issues and generating AI-assisted repairs. Furthermore, the ability to manually control MCP server startup addresses performance and resource management concerns for developers using multiple AI tools.
For enterprise environments, the ability to enforce default models is a significant step toward consistent AI-assisted development standards. It allows organizations to balance developer flexibility with the need to steer teams toward approved or tested models.
The integration of diagnostic 'Fix' actions directly into the IDE's intention menu represents a shift toward more proactive AI assistance. By reducing the number of steps required to move from an error report to a proposed solution, the update aims to keep developers within their flow state.
The manual control over MCP server startup is a practical improvement for power users who manage multiple AI integrations. By preventing automatic server initialization, developers can avoid unnecessary overhead and potential conflicts between different AI tools running within the same IDE environment.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
Users should monitor the stability of the new agent-mode diagnostic features and the impact of the mandatory IDE version upgrade. It remains unknown how these specific model-default settings will interact with future, more complex agentic workflows or if additional model providers will be integrated into the managed settings interface in subsequent updates.
The requirement for JetBrains IDE 2025.2 or later may cause temporary disruption for teams unable to update their IDE environment immediately. Users should verify their current IDE version before attempting to update the plugin.
The effectiveness of the 'Fix' action in agent mode versus ask mode will depend on the specific context of the code error. It is currently unknown how the plugin handles edge cases where the AI might suggest incorrect or insecure code patches during these automated diagnostic sessions.
Future updates may clarify how these managed settings interact with custom-trained models or third-party providers, as the current documentation focuses on standard Copilot agent models.