AI News Summary - July 12, 2026

Claude Code Introduces Multi-Repository Context Windows

Anthropic expanded Claude Code's capabilities to handle context across up to 10 linked repositories simultaneously, with a combined context window of 500K tokens. The update includes intelligent dependency mapping that automatically identifies and includes relevant code from connected projects. Early access users report 40% faster development time when working on microservices architectures.

MCP 2.0 Specification Released with Streaming Support

The Model Context Protocol consortium published version 2.0, adding native support for streaming data sources and real-time context updates. The new specification enables AI assistants to maintain persistent connections to databases, APIs, and file systems with automatic change detection. Over 50 tool providers have committed to implementing MCP 2.0 within the next quarter.

Open-Source Agent Framework "Orchestrate" Reaches 1.0

The collaborative agent framework Orchestrate hit its stable 1.0 release after 18 months of development, featuring built-in support for multi-agent workflows and human-in-the-loop approvals. The framework includes pre-built integrations for all major LLM providers and a visual workflow designer for non-technical users. The project has already accumulated 23K GitHub stars and is being used in production by several Fortune 500 companies.

GitHub Copilot Workspace Adds Automated Code Review Agents

GitHub announced AI-powered code review agents within Copilot Workspace that can analyze pull requests for security vulnerabilities, performance issues, and style consistency. The agents use a combination of static analysis and LLM reasoning to provide contextual feedback with suggested fixes. The feature is rolling out to GitHub Enterprise customers this week with general availability planned for August.