Add git-ai integration to quantify roborev’s impact on AI-generated code

[beettlle]

Summary

Support integration with git-ai so that AI-generated code produced during roborev workflows—particularly when roborev fix and roborev refine invoke agents to write or modify code—is tracked via git-ai’s authorship metadata. This would allow teams and organizations to measure roborev’s contribution to their AI-assisted development.

Motivation

Organizations increasingly use git-ai to track AI-generated code and measure AI impact at commit, PR, repository, and organization levels. Teams using both roborev and git-ai today cannot attribute roborev-driven fixes and refinements to roborev. As a result, roborev’s impact on code quality and AI utilization is invisible in dashboards, compliance reports, and cost analyses.

Why this matters

• Visibility into roborev’s impact — When roborev invokes agents to fix or refine code, that output should be measurable. git-ai integration would let users see how much AI-generated code came from roborev workflows vs. direct agent use.
• Enterprise metrics — Organizations using git-ai for AI usage, attribution, and governance would get a more complete picture of AI adoption, including roborev’s role.
• Consistency with existing tooling — git-ai is vendor-agnostic and integrates with many of the same agents roborev uses (Cursor, Claude Code, Gemini, Copilot, OpenCode, Droid). Integration would fit into existing AI-tracking workflows.
• Value of continuous review — roborev’s continuous review model becomes more compelling when the AI-generated fixes it drives can be tracked and attributed. git-ai integration would close that gap.

Use case

A team uses roborev for post-commit review and roborev fix to address findings. Today, git-ai only tracks code from agents that call it directly. Code produced by agents invoked by roborev during fix or refine is not attributed to roborev. With integration, that code would be attributed (e.g., to roborev + the underlying agent), so teams can see how much AI-generated code comes from roborev-driven fixes.

Requested behavior

Support git-ai authorship and prompt tracking for code generated by roborev workflows (e.g., roborev fix, roborev refine, and any future flows that invoke agents to modify code), so that roborev-driven edits appear in git-ai’s metrics and AI blame without requiring changes to downstream consumers.

References

• git-ai — Git extension for tracking AI-generated code
• git-ai: Add your Agent — Integration guide for coding agents
• Git AI Standard v3.0.0 — Authorship log format