Coding Tools
GitHub Copilot July Update for Visual Studio
GitHub Changeloghigh confidence
Why it matters
- →The new agent improves task efficiency with fewer interactions.
- →Built-in skills from .NET and Azure teams enhance workflow customization.
- →Organization-level instructions streamline team-wide Copilot responses.
GitHub has released a July update for Copilot in Visual Studio, featuring a new agent option in public preview. Built on the Copilot SDK, this agent aims to improve task completion with less interaction. The update also includes built-in skills from the .NET and Azure teams, and a feature for reviewing selected code with inline comments. Organization-level custom instructions are now available, enhancing team-wide customization. Some features require specific Copilot plans, but the update overall enhances Copilot's adaptability and efficiency.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1GitHub Copilot May Update for Visual Studio 2026
GitHub Changelog · June 4, 2026 · Same story
- 2GitHub Copilot June Update for Visual Studio
GitHub Changelog · July 15, 2026 · Same story
Follow this story
Open the full story →GitHub Copilot Enhances Code Review with Agent Skills
More in Coding Tools
Coding Toolscoding
Claude Code v2.1.316: Subagents and Gateway Budgets
Anthropic is tightening control over Claude Code's autonomous behavior with new environment variables that let you disable 'ultrathink' or cap subagent effort levels. For teams, the introduction of pooled group budgets on the gateway allows shared monthly spending limits across members, a practical fix for enterprise cost management. The release also patches critical bugs where shell commands would hang or crash on exit, stabilizing the developer experience significantly.
Claude Code Releases
Coding Toolscoding
llama.cpp b12022 adds ROCm 10 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally bringing ROCm 10.0 to Linux and Windows alongside CUDA 13.4, effectively closing the hardware gap for AMD users who were previously left behind. The inclusion of native Snapdragon support on Linux marks a significant step toward running local LLMs on ARM-based mobile chips without emulation overhead. While KleidiAI on Apple Silicon is disabled in this build, the broader expansion to diverse accelerators ensures that developers aren't locked into NVIDIA ecosystems anymore.