Adoption is broad. Many contributors are committing agent-authored work, which is what a real organisational rollout looks like.
Assisted or autonomous
Assisted — a developer’s identity authored the commit and the agent added itself as co-author. The agent ran git commit during a working session, with a person there. Autonomous — a bot identity is the author, no human in the loop. Across the whole index, 93.7% of agent commits are assisted.
Agent share by month
Peak month: 12.5%. Bars are scaled to that peak, not to 100%.
Which agents
| Claude Code | Anthropic | 151 | 4% |
| Cursor Agent | Anysphere | 99 | 2.6% |
| GitHub Copilot | GitHub | 2 | 0.1% |
| Devin | Cognition | 1 | 0% |
Verify it yourself
Sample commits that matched. Every one links to GitHub — check the trailers.
70a37b98Reset the Codex failure gate when a successful fetch is withheldClaude Code97fc3b45Clear a stale Codex connectivity failure when publication is withheldClaude Coded8d03a12Price OpenCodex usage once per entry and stop per-entry catalog/overlay reClaude Codeb4448cdcDocument why deferred file starvation via growth cannot occurCursor Agent35ff9a18Sort deferred files by remaining bytes and record scan attemptsCursor Agentcb1a5d6aAddress review feedback: fix dedup state leak and test determinismCursor Agent573652aei18n: improve Simplified Chinese localization (#1059)GitHub Copilotec2005bdAdd Copilot multi-account support (#637)GitHub Copilot72aaaf3dfix: preserve Devin one-percent usage (#1938)Devin
What was excluded
| Commits in window | 4,489 |
| Merge commits removed | −713 |
| CI / dependency bots removed | −0 |
| Considered | 3,776 |
Bots are excluded from the denominator entirely. Counting dependabot as “AI-written code” is the fastest way to produce a meaningless number.
How this was measured
Measured over February–August 2026, cloned directly. Detection is deterministic string matching against signatures agents write about themselves — no heuristics, no style analysis. Merge commits and CI bots are excluded from the denominator. The figure is a floor: squash merges strip trailers, so real usage is higher. Contributor counts are aggregate — no individual is named.
Run it yourself: npx qmmit-cli steipete/CodexBar