AI News Summary - November 14, 2026

Claude Code Introduces Multi-Repository Context Awareness Anthropic has released Claude Code 3.5, which can now maintain context across multiple linked repositories simultaneously. The update allows developers to refactor code that spans microservices architectures while automatically tracking dependencies. Early adopters report 40% faster completion times for complex cross-service changes.

MCP Specification 2.0 Adds Streaming Tool Capabilities The Model Context Protocol consortium announced version 2.0, introducing bidirectional streaming for long-running tool executions and stateful context sessions. Major providers including OpenAI, Anthropic, and Google have committed to implementation by Q1 2027. The update enables AI agents to provide real-time progress updates during database migrations, build processes, and other extended operations.

Open-Source Agent Framework "Nexus" Reaches Production Status The Nexus autonomous agent framework has hit version 1.0 after 18 months of development, offering built-in safeguards and resource limits for production deployments. Nexus supports Python, TypeScript, and Rust, with native MCP integration and a plugin system for custom tools. The framework has already been adopted by 200+ companies for internal automation workflows.

GitHub Copilot Workspace Adds AI-Powered Code Review GitHub announced Copilot Workspace now includes an AI reviewer that analyzes pull requests for security vulnerabilities, performance bottlenecks, and architectural inconsistencies before human review. The system integrates with existing CI/CD pipelines and can automatically suggest fixes for common issues. Beta testing showed a 60% reduction in bugs reaching production.