AI News Summary - June 15, 2026

GitHub Copilot Workspace Adds Multi-Repository Context

GitHub announced Copilot Workspace now supports simultaneous context from up to 10 repositories, enabling developers to refactor code across microservices architectures. The feature uses an enhanced MCP implementation that maintains consistent context windows while reducing token costs by 40%. Early access begins next week for Enterprise customers.

Anthropic Releases Claude Code 3.5 with Extended Test Generation

Claude Code 3.5 now automatically generates comprehensive test suites including edge cases, integration tests, and performance benchmarks. The update introduces "Test-Driven Assistance" mode where the AI writes tests first, then implements code to pass them. Anthropic reports 35% fewer production bugs in internal trials compared to previous versions.

OpenMCP Framework Reaches 1.0 Milestone

The open-source OpenMCP project hit version 1.0, establishing a standardized protocol for AI context sharing across 50+ IDEs and development tools. Major adoptions include JetBrains Fleet, Cursor, and Zed, enabling seamless context handoffs between different coding assistants. The framework now supports real-time collaborative coding sessions with persistent context across multiple AI providers.

Aider Introduces Voice-Controlled Code Review

The popular open-source pair programming tool Aider added natural voice commands for code review workflows. Developers can now verbally approve changes, request modifications, or ask clarifying questions while the AI maintains full codebase context. The feature integrates with MCP to preserve conversation history across sessions.