Coding Tools
GitHub Copilot August Update Enhances Control
GitHub Changeloghigh confidence
Why it matters
- →Developers gain more control over AI model reasoning and usage.
- →Organization-level custom agents enhance team collaboration and efficiency.
- →Adjustable thinking effort allows for tailored AI assistance, optimizing resource use.
GitHub has released an August update for Copilot in Visual Studio 2026, enhancing control over AI model usage and reasoning. Key features include organization-level custom agents, adjustable model thinking effort, and improved model management. Users can now tailor AI assistance to their specific needs, whether for simple tasks or complex debugging. This update is available across all GitHub Copilot plans, aiming to provide a more personalized and efficient coding experience.
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 App Now Supports Any Model Provider
Matt Wolfe · June 10, 2026 · Same story
- 2GitHub Copilot Enhancements in VS Code June 2026
GitHub Changelog · July 8, 2026 · Same story
- 3GitHub Copilot July Update for Visual Studio
GitHub Changelog · July 30, 2026 · Same story
Follow this story
Open the full story →GitHub Copilot Enhances Customization and Performance
3 developments
- Aug 28 · GitHub Changelog
GitHub Copilot Enhances Customization and Performance - Aug 28 · GitHub Changelog
GitHub Copilot August Update Enhances Control (This article)↳ The August update adds adjustable thinking effort, organization-level custom agents, and improved model management features. - Aug 31 · GitHub Changelog
GitHub Copilot Enhancements in VS Code August 2026
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.
llama.cpp Releases
Coding Toolscoding
llama.cpp b12023 fixes shared prefix bug
This release addresses a specific inference artifact where decision prefixes were incorrectly bleeding into output embeddings. By capping shared prefixes at the first output position, llama.cpp ensures that option tokens remain distinct for parent and child tasks. This is a targeted fix for multi-step reasoning or voting workflows rather than a performance boost. It stabilizes complex generation patterns without changing the underlying model architecture.
llama.cpp Releases