AI News Summary - June 27, 2026

Claude Code Introduces Multi-Repository Context Awareness

Anthropic announced Claude Code now supports simultaneous context across up to 10 linked repositories, enabling developers to refactor code across microservices architectures seamlessly. The update includes a new "repository graph" visualization that maps dependencies and suggests related files during coding sessions. Early access users report 40% faster cross-service refactoring workflows.

MCP 2.0 Specification Released with Bidirectional Tool Calling

The Model Context Protocol consortium published MCP 2.0, introducing bidirectional tool invocation where AI agents can now expose their own capabilities as MCP tools for other agents to consume. This enables unprecedented agent-to-agent collaboration without custom integration code. Over 200 organizations have committed to implementing the new spec within Q3 2026.

Open-Source AgentFlow Framework Reaches 50K GitHub Stars

AgentFlow, the declarative agent orchestration framework, hit a major milestone with its v3.0 release featuring visual workflow debugging and automatic failure recovery. The framework now supports 15 LLM providers out-of-the-box and includes pre-built templates for common agent patterns like research, data analysis, and customer support. The project has seen contributions from over 800 developers worldwide.

Cursor Launches "Team Memory" for Collaborative AI Context

Cursor IDE unveiled Team Memory, allowing AI coding assistants to learn from collective team decisions, code patterns, and architectural choices across all team members. The feature uses privacy-preserving federated learning to build shared context without exposing individual codebases. Beta testing shows 30% reduction in repetitive code review comments.