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: 24.3%. Bars are scaled to that peak, not to 100%.
Which agents
| Claude Code | Anthropic | 105 | 9.3% |
| Cursor Agent | Anysphere | 20 | 1.8% |
| GitHub Copilot | GitHub | 8 | 0.7% |
Verify it yourself
Sample commits that matched. Every one links to GitHub — check the trailers.
67e16a9erefactor(dotcom): stop declaring the legacy file owner columns (#10653)Claude Codeea63b7ddrefactor(dotcom): remove legacy user-owns-file code paths (#10391)Claude Coded9c7b429docs: fix tldraw.dev page errors caught by the dotdev smoke test (#10539)Claude Code4352e0eafeat(editor): add tleditors global registry for mounted editors (#9904)Cursor Agent0714aa9afix(editor): keep the tick loop alive when a listener throws (#9946)Cursor Agentb518d58atest(driver): carry held modifiers on simulated keyboard events (#9406)Cursor Agent58c64502docs(skills): consolidate voice guides and point skills at code (#9844)GitHub Copilot0527a7d5feat: commenting followups (#9782)GitHub Copilot48a9e773feat: attribute duplicated and pasted notes to the current user (#9566)GitHub Copilot
What was excluded
| Commits in window | 1,187 |
| Merge commits removed | −0 |
| CI / dependency bots removed | −61 |
| Considered | 1,126 |
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 tldraw/tldraw