GitHub has announced a transformative expansion of Copilot’s capabilities, marking a significant evolution in AI-assisted development that promises to reshape DevOps practices. As part of Microsoft’s 50th-anniversary celebration, GitHub is rolling out Agent Mode with Model Context Protocol (MCP) support to all Visual Studio Code users, transitioning Copilot from a code completion tool to a fully agentic development partner.
Agent Mode: Beyond Code Completion
Agent Mode is the most revolutionary aspect of this update, fundamentally changing how developers interact with their tools. Unlike traditional chat or multi-file edits, Agent Mode can independently translate ideas into code, automatically identifying necessary subtasks and executing them across multiple files.
For DevOps professionals, Copilot can now handle complex infrastructure tasks, suggest terminal commands, make tool calls and even self-heal runtime errors. The system achieves an impressive 56% pass rate on SWE-bench Verified with Claude 3.7 Sonnet, and its capabilities are expected to grow as reasoning models advance.
Multi-Model Support and Premium Requests
GitHub has also introduced multi-model support, making Anthropic Claude 3.5, 3.7 Sonnet, Google Gemini 2.0 Flash and OpenAI models generally available through a new premium request system. This gives DevOps teams unprecedented flexibility to select the right model for specific tasks.
Paid plans receive varying allotments of premium requests (300 for Pro, 300 for Business, and 1,000 for Enterprise) while maintaining unlimited access to the base model (currently OpenAI GPT-4o). A new Pro+ tier offers 1,500 monthly premium requests for $39/month, and organizations can opt into pay-as-you-go for additional requests at $0.04 each.
MCP: The “USB Port for Intelligence”
Perhaps most significant for infrastructure and operations teams is the public preview of the Model Context Protocol (MCP). This protocol is a “USB port for intelligence,” allowing Agent Mode to access any context or capabilities needed.
The new open-source GitHub MCP server enables teams to add GitHub functionality to any LLM tool supporting MCP. This means DevOps workflows can now benefit from agents that understand database schemas, query telemetry and manage infrastructure with contextual awareness.
Transforming DevOps Workflows
For DevOps practitioners, these advancements represent a fundamental shift in how infrastructure can be managed:
- Infrastructure as Code Acceleration: Agent Mode can analyze existing infrastructure configurations, suggest improvements and implement them across multiple files.
- Troubleshooting Enhancements: With access to logs and system state via MCP, Copilot can help diagnose issues and suggest remediation steps.
- Workflow Automation: Complex tasks like “Update the CI pipeline to include the new security scan step” can be decomposed and implemented automatically.
- Cross-Platform Integration: MCP allows Copilot to interface with various tools in the DevOps stack, creating a unified experience.
The GitHub local MCP server enhances platform integration, enabling capabilities like searching repositories, managing issues and creating PRs directly from VS Code.
Mitch Ashley, VP and Practice Lead, DevOps and Application Development, The Futurum Group, states, “We are witnessing software development shift from augmented development to agent-based, or agentic, developed software. GitHub shifted Agent mode from a preview into its rollout phase in less than a month. We are at the forefront of increased developer productivity and software delivery velocity.”
Looking Ahead
As these capabilities mature, we can expect DevOps workflows to become increasingly automated, with human expertise focused on strategic decisions rather than implementation details. The combination of agentic capabilities and multi-model flexibility provides a foundation for the next generation of infrastructure management.
GitHub’s progression from code completion to agentic development assistant marks a significant milestone in the AI-driven transformation of software development. For DevOps professionals, these tools offer productivity enhancements and a fundamental reimagining of how infrastructure and operations can be managed in collaborative human-AI partnerships.
