GitHub Updates Usage Metrics API with Custom Skills and MCP Servers
GitHub has updated its Copilot usage metrics to display specific names for organization-created skills, custom agents, Model Context Protocol (MCP) servers, slash commands, and plugins instead of grouping them broadly.
Radwan AbdulhadiEditor and technology writer
Enhanced Visibility in Copilot Usage Metrics
According to the [GitHub Changelog](https://github.blog/changelog/2026-10-22-custom-skills-agents-and-mcp-servers-in-usage-metrics-api/), GitHub Copilot usage metrics now show the names of the skills, custom agents, Model Context Protocol (MCP) servers, slash commands, and plugins that an organization creates. Rather than grouping these custom items under generic labels like other or custom, the reporting provides clearer visibility.
This update directly builds upon the agentic CLI customization metrics that were initially released by the platform in September.
Detailed Breakdown of Named Entries and Arrays
Within enterprise and organization reports—including per-user 1-day and 28-day reports, aggregated 1-day reports, and day_totals in aggregated 28-day reports—several arrays now display the configured name of a customer-defined item instead of generalized groupings.
The affected fields include totals_by_skill, totals_by_custom_agent, totals_by_mcp, totals_by_slash_cmd, and totals_by_plugin. Furthermore, each list now accommodates up to 25 entries per record, expanding significantly from the previous limit of five entries.
By increasing the entry limit, administrators can view a broader segment of the long tail of customizations utilized by their developers. Any remaining activity that falls outside these top listings is combined into a trailing other entry, while interaction counts and distinct usage counts continue to be calculated according to previous methods.
Tracking Adoption and Internal Enablement
The customizations that matter most to an organization are frequently those built by its own internal teams. With these enhancements, administrators gain the capability to see which internal skills, custom agents, and MCP servers developers actively use.
This granular insight allows teams to track adoption rates accurately after rolling out a new tool and to focus organizational enablement efforts precisely where they are needed most.
Coverage Scope Across Integrations and Applications
Names are reported across multiple environments, including the Copilot CLI in both interactive sessions and prompt mode, the GitHub Copilot app, Agent Client Protocol (ACP) integrations such as JetBrains, and the Copilot SDK.
To support this functionality, prompt mode, the mobile or desktop app, and ACP integrations require Copilot CLI version 1.0.97 or later. Meanwhile, the Copilot SDK requires version v1.0.21 or later, and integrated development environments report names when their agent experiences run on these compatible SDK versions. However, names are not reported from an IDE's built-in chat telemetry.
Configuration Standards, Accuracy, and Access
Configured names are reported as they are defined, though they are trimmed and restricted to a maximum of 128 characters. Names that resemble file paths remain grouped under other or custom, and administrators are advised to review names before sharing reports externally.
Because these metrics originate from client telemetry, they may occasionally undercount activity due to outdated clients or blocked telemetry. Consequently, GitHub notes that these metrics are best utilized for understanding general trends and adoption rather than exact accounting.
Historical data reflects that names appear beginning with report days processed after the release, while earlier days retain the previous grouping structure. Reports remain accessible to enterprise owners, billing managers, organization owners, and any user assigned a custom organization or enterprise role with the "View Copilot Metrics" permission, provided the usage metrics policy is enabled.
Developers and administrators can consult the [Copilot usage metrics API documentation](https://docs.github.com/rest/copilot/copilot-usage-metrics) to get started with implementing these telemetry reporting capabilities.
Sources
- GitHub ChangelogCustom skills, agents, and MCP servers in usage metrics API
Continue chronologically
Related entity coverage
Related stories
GitHub Adds BYOK Support to Copilot Usage API
GitHub updates its Copilot usage metrics API to separate bring-your-own-key usage and enterprise custom models from Copilot-hosted data.
GitHub adds CLI customization metrics to Copilot API
GitHub has expanded Copilot’s usage metrics API with activity data for CLI skills, custom agents, MCP servers, slash commands and plugins.
GitHub Copilot Integrates with JetBrains Air
GitHub Copilot is now available in JetBrains Air, enabling developers to delegate coding tasks, guide execution, and review results locally or in the cloud.
GitHub Intelligence Empowers AI Agents with Team Context
GitHub introduces GitHub intelligence, using code, context, and development history to help AI agents adapt to team workflows and conventions.
