Add GitHub extractor for repositories, code ownership, and CI/CD

[ravisuhag]

Context

Code-to-data relationships are among the most commonly needed by AI agents. An agent debugging a data issue needs to know what code produces a table, when it last changed, and what the deploy pipeline looks like.

Scope

New GitHub extractor that captures:

• Repositories — name, description, language, topics, visibility
• Code ownership — CODEOWNERS file parsing, team-to-path mappings
• CI/CD pipelines — GitHub Actions workflows, their triggers, and what they deploy
• Pull requests — recent merges that changed data-related code (dbt models, migrations, pipeline configs)
• Relationships — repo → assets it produces (via convention or config), team → repo ownership

Approach

• Use GitHub API (REST or GraphQL) for metadata extraction
• Support organization-wide scanning or per-repo configuration
• Emit typed relationships linking repos/code to data assets where inferrable
• Rate limiting and pagination handling for large organizations

Why

GitHub is likely the highest-value new extractor because it connects code to data assets — the relationship most commonly asked about by AI agents. "What code produces this table?" and "when did this pipeline last change?" are foundational questions.

References

• Meteor Roadmap — Expand Coverage