GitHub Copilot
The GitHub Copilot connector imports Copilot usage metrics for your organisation and teams, so your AI Coding dashboard shows how Copilot is adopted and used across the business — code suggestions, chat activity, pull request summaries and active-user trends.
Authentication
GitHub Copilot is not a standalone connector. It is enabled through your existing GitHub or GitHub App connector by turning on AI usage data in the connector's settings after installation. Logilica reuses the same GitHub authentication — OAuth, a Personal Access Token or a GitHub App installation — to read Copilot metrics for your organisation.
See Connecting Tools for where to enable AI usage data on the GitHub connector.
What's Imported
Logilica first reads your Copilot seat assignments to determine the granularity of import:
If seats are assigned through teams, Logilica imports metrics for each team that has Copilot seats.
If no team assignments are found, Logilica imports metrics for the whole organisation.
For each organisation or team, and for each day, Logilica imports:
Code completions — accepted suggestions, total suggestions, and lines added versus lines suggested by AI.
Interaction types — IDE and dotcom chat, chat insertions, chat copies, and pull request summaries created by Copilot, plus a combined interaction count.
Most-used model — the model with the highest usage across completions, chat and pull request summaries.
Active-user percentage — the share of seat holders who were active that day.
The Copilot metrics API returns aggregated data only — there is no per-user breakdown. Logilica records usage against the organisation or the team as a group, not against individual developers. To see how active a group is day to day, use the active-user percentage rather than per-person figures.
Good to Know
Kept in sync automatically. After the initial import, Logilica refreshes this connection on a regular schedule — you don't need to re-trigger imports.
The import window is the last 100 days — not the 180 days used by Cursor and Claude Code, and not the 6-month window used by other connectors. This is a limit of the Copilot metrics API.
Data is aggregate-only. Because the Copilot API does not expose individual usage, per-developer breakdowns are not available for Copilot, even though they are available for Cursor and Claude Code.
If no Copilot seats are assigned, nothing is imported and the connector finishes without adding data.
GitHub Copilot data feeds the AI Coding dashboard, where AI usage is shown alongside your delivery metrics.
Last updated