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: 33.2%. Bars are scaled to that peak, not to 100%.
Which agents
| GitHub Copilot | GitHub | 643 | 22.6% |
| Claude Code | Anthropic | 5 | 0.2% |
| Cursor Agent | Anysphere | 3 | 0.1% |
Verify it yourself
Sample commits that matched. Every one links to GitHub — check the trailers.
2e953c46Fix broken relative links in integration-test-analysis SKILL.md (#85293)GitHub Copilot0a1187caRemove team specific ExternalAccess assemblies (#85135)GitHub Copilot4af55b27Make CI test payload artifacts retry-safe (#85248)GitHub Copilot20d29497Let F# take part in the Navigate To search that runs while the solution loClaude Codee47c73cbLet F# create a workspace project context without blocking the main threadClaude Code9a38b298Embed F# external access resources under the name their accessor asks for Claude Code74d07c07Tighten fuzzy edit-distance threshold for length-5 patterns (#82459)Cursor Agenta80bb709Improve nav-to fuzzy pre-filtering with q-gram bigram bitset and correctedCursor Agent31befaa8Improve navigate-to filtering to avoid scanning documents unnecessarily. (Cursor Agent
What was excluded
| Commits in window | 3,581 |
| Merge commits removed | −628 |
| CI / dependency bots removed | −112 |
| Considered | 2,841 |
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 dotnet/roslyn