of 1,496 commits were committed by an AI agent — a floor, not an estimate

5% agent · top contributor 25.3%32 contributors

org-wide

Adoption is broad. Many contributors are committing agent-authored work, which is what a real organisational rollout looks like.

agent commits

1,421

human commits

agent contributors

25.3%

largest contributor

Assisted or autonomous

75assisted

0autonomous

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

Jan26

Feb26

Mar26

Apr26

May26

Jun26

Jul26

Aug26

Sep26

Peak month: 11%. Bars are scaled to that peak, not to 100%.

Which agents

Cursor Agent	Anysphere	30	2%
GitHub Copilot	GitHub	25	1.7%
Claude Code	Anthropic	21	1.4%
Devin	Cognition	1	0.1%

Verify it yourself

Sample commits that matched. Every one links to GitHub — check the trailers.

692b22c5fix(drizzle): export generated pgSchema for drizzle-kit (#10770)Cursor Agent
8a3fa826docs: make AI chat failures recoverable (#10737)Cursor Agent
a7962147feat(oauth-provider): remove silenceWarnings config and well-known endpoinCursor Agent
5cb6a0d8docs: explain the 1.7 account schema transition (#11196)GitHub Copilot
882cf9e5fix(admin): use authoritative session reads for authorization (#10187)GitHub Copilot
90d509e0fix(adapter): fail closed on update misses (#10180)GitHub Copilot
1d9b55cffix(cimd): answer the socket's all-addresses lookup in the Node transport Claude Code
bb8d7c45fix(oauth-provider): backport CIMD grant intersection (#11010)Claude Code
a7962147feat(oauth-provider): remove silenceWarnings config and well-known endpoinClaude Code
51476fe9docs: fix authClient import in Auth0 and Clerk migration guides (#11248)Devin

What was excluded

Commits in window	1,665
Merge commits removed	−56
CI / dependency bots removed	−113
Considered	1,496

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 better-auth/better-auth

Scan another repository

Try