Coding Tools
GitHub Copilot metrics now name custom agents and MCPs
GitHub Changeloghigh confidence
Why it matters
- →Enables precise tracking of internal AI tool adoption beyond generic usage stats.
- →Expands visibility into the 'long tail' of custom integrations with a 25-entry limit.
- →Critical for enterprise governance and ROI analysis of custom agent deployments.
GitHub has updated its Copilot usage metrics API to display specific names for custom skills, agents, and Model Context Protocol (MCP) servers, replacing previous generic groupings like 'other' or 'custom'. The update expands the number of visible entries per record from five to twenty-five, allowing organizations to track adoption of niche internal tools more accurately. These changes apply to enterprise and organization reports across CLI, app, and SDK integrations, provided clients are updated to version 1.0.97 (CLI) or v1.0.21 (SDK). Administrators can now monitor which custom AI workflows are actually being used by developers rather than relying on aggregated totals.
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 adds VS Code Agents metrics
GitHub Changelog · September 11, 2026 · Same story
- 2GitHub Copilot CLI adds agentic usage metrics
GitHub Changelog · September 17, 2026 · Same story
Follow this story
Open the full story →GitHub Copilot metrics now name custom agents and MCPs
3 developments
- Oct 22 · GitHub Changelog
GitHub Copilot metrics now name custom agents and MCPs (This article) - Oct 23 · GitHub Changelog
Copilot API now tracks non-IDE tool usage↳ Copilot API tracks non-IDE tool usage via totals_by_client to expose external agent traffic - Oct 27 · GitHub Changelog
GitHub Copilot adds server-to-server auth for orgs↳ Copilot adds server-to-server auth for orgs allowing GitHub Apps to authenticate directly via installation tokens independent of human login
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