Skip to main content

AI Usage > Repository

Written by Tom Williams

Dataset: AI Usage

Entity: AI Usage Measurement

Field ID: repository_full_name

Type: Text

Description: The full name of the repository the measurement relates to, in owner/name form.

Note: Only populated on REPOSITORY records, which only GitHub Copilot publishes.

Source: App

Transformation logic: N/A

App Mapping

GitHub Copilot

the repository owner and name of the per-repository report

Anthropic Claude (coming soon)

N/A - attributes no usage to a repository

Cursor (coming soon)

N/A

OpenAI Platform (coming soon)

N/A

OpenAI Codex (coming soon)

N/A

Reporting Use Cases

The Repository field is the only way to see AI impact per codebase. It is the natural join point between AI usage and the pull request data in the Issues & Pull Requests dataset.

  • AI impact per codebase: Filter to aggregation_level = REPOSITORY and group by repository with SUM(pull_requests_created) to see where AI is producing merged work.

  • Comparing with total activity: Repository records carry only the AI-attributed pull request counts. Compare them against the equivalent totals in Issues & Pull Requests to get an AI share of all pull requests.

  • GitHub Copilot only: No other provider attributes usage to a repository, so a repository report covers Copilot alone.

  • Repositories with no AI activity are absent: A repository with no Copilot pull request activity produces no record at all, rather than a record of zeroes.

Did this answer your question?