AI News Summary - April 30, 2026

Claude Code Introduces Multi-Repository Analysis

Anthropic's Claude Code now supports simultaneous analysis across up to 50 linked repositories, enabling developers to refactor code and track dependencies across entire microservices architectures. The update includes a new "architectural understanding" mode that builds context graphs of how services interact. Early adopters report 40% faster debugging in complex distributed systems.

MCP Gateway Protocol Reaches 1.0 Release

The Model Context Protocol community announced MCP Gateway 1.0, a standardized proxy layer that allows AI agents to securely access multiple MCP servers through a single authenticated connection. The specification includes built-in rate limiting, audit logging, and permission scoping for enterprise deployments. Over 200 organizations contributed to the final specification during the six-month public review period.

OpenHands Agent Framework Adds Visual UI Testing

The open-source OpenHands project released v3.0 with native support for visual regression testing and browser automation workflows. Agents can now capture screenshots, compare visual differences, and autonomously fix CSS/layout issues across multiple viewport sizes. The framework has been integrated into several CI/CD platforms with pre-built GitHub Actions.

Cursor Launches Collaborative Agent Workspaces

Cursor unveiled team workspaces where multiple AI agents can collaborate on the same codebase with different specialized roles (architect, reviewer, tester). The feature uses a novel "agent communication protocol" that prevents conflicting edits while allowing parallel work streams. Beta teams report 3x faster feature development when using specialized agent teams versus single general-purpose assistants.