AI News Summary - July 7, 2026

Claude Code Introduces Multi-Repository Context Awareness

Anthropic released Claude Code 3.5, featuring breakthrough multi-repository understanding that maintains context across up to 15 linked codebases simultaneously. The update reduces hallucinations in monorepo environments by 47% and includes automatic dependency graph visualization. Early access users report significant improvements in microservices development workflows.

MCP Gateway Protocol Reaches 1.0 Release

The Model Context Protocol consortium announced MCP Gateway 1.0, a standardized routing layer that enables seamless context sharing between different AI providers. Major adopters include Cursor, JetBrains, and VS Code, with the protocol now supporting real-time collaborative context streams and encrypted credential management. Over 200 MCP servers are now publicly available in the official registry.

AutoDev Framework Adds Visual UI Testing Agents

The open-source AutoDev project launched agent capabilities for autonomous visual regression testing, combining screenshot analysis with code-level understanding. The framework can now identify UI bugs, generate Playwright tests, and automatically open PRs with fixes across web and mobile applications. Initial benchmarks show 73% accuracy in detecting layout regressions without human intervention.

OpenHands Releases Multimodal Agent SDK

OpenHands unveiled a production-ready SDK for building multimodal AI agents that combine vision, code execution, and browser automation in a single pipeline. The toolkit includes pre-built templates for data analysis, document processing, and web scraping tasks, with native support for tool calling across Claude, GPT-4, and open models. Over 5,000 developers joined the beta in the first 24 hours.