Copilot Chat v0.33.2025110403: Build & API Fixes
Copilot Chat v0.33.2025110403 fixes build blockers, cloud agent detection, and adds Anthropic web search settings for BYOK providers.
copilot
Updates from GitHub Copilot ecosystem
Copilot Chat v0.33.2025110401: Subagent & Web Search
Subagent2 lands in plan mode, Anthropic web search gets proper citations, and experimental GPT-5 prompts are removed. Security fixes included.
Copilot Chat v0.33.2025110305: CLI Edits & Git Context
Copilot Chat v0.33.2025110305 brings CLI edit persistence, Git context for commits, and worktree support. Internal refactoring improves session handling and code quality.
GitHub Game Off 2025: Build a Game in November, Theme is WAVES
GitHub's annual Game Off jam runs through November with the theme WAVES. Build a game in any language, push to a public repo, submit by December 1. AI-assisted dev is allowed. Perfect for first-timers or veterans.
Copilot Chat v0.33.2025103102: Memory Tool & Edit Fixes
Copilot Chat adds Memory tool support, fixes parallel edit conflicts, and improves CLI integration with better context provider handling.
Copilot Chat v0.33.2025103101: CLI & Model Fixes
Copilot Chat v0.33.2025103101 refactors CLI loading, fixes thinking content in custom models, and adds cloud agent file tracking. Stability and compatibility improvements across the board.
How GitHub Tests MCP Server Quality: Offline Evaluation Deep Dive
GitHub built an automated pipeline to test how well LLMs select MCP tools and supply correct arguments. They treat tool selection as multi-class classification and track four argument-quality metrics to catch regressions before users see them.