AI News Summary - April 29, 2026

Claude Code Introduces Multi-Repository Context Awareness

Anthropic's Claude Code now supports simultaneous context from up to 10 linked repositories, addressing a major pain point for developers working on microservices architectures. The update includes intelligent dependency tracking that automatically pulls relevant code from connected repos when generating suggestions. Early adopters report 40% fewer context-switching errors in cross-service refactoring tasks.

MCP 2.0 Specification Released with Bidirectional Tool Calling

The Model Context Protocol consortium published MCP 2.0, introducing bidirectional tool invocation that allows both clients and servers to initiate actions. This enables more sophisticated agentic workflows where tools can proactively request additional context or user input during long-running operations. Over 200 MCP-compatible tools have already committed to supporting the new specification.

Open-Source AgentGraph Framework Reaches Production Stability

AgentGraph 1.0, an open-source framework for building multi-agent systems with visual workflow graphs, officially launched after 8 months in beta. The framework features built-in observability, automatic retry logic, and native support for popular LLM providers including OpenAI, Anthropic, and local models via Ollama. The project has gained 15K GitHub stars and is already deployed in several enterprise environments.

GitHub Copilot Workspace Adds AI-Powered Code Review

GitHub announced Copilot Workspace now includes automated PR reviews that go beyond static analysis to understand business logic and architectural patterns. The feature uses repository-wide context to identify potential issues with data flow, security implications, and consistency with existing patterns. The review system integrates directly into the PR workflow and can be customized with team-specific guidelines.