← the index
microsoft/onnxruntime
29.3%
of 1,428 commits were committed by an AI agent — a floor, not an estimate
29.3% agent · top contributor 25.6%66 contributors
org-wide

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

418
agent commits
1,010
human commits
66
agent contributors
25.6%
largest contributor

Assisted or autonomous

417assisted
1autonomous

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

Feb26
Mar26
Apr26
May26
Jun26
Jul26
Aug26
Sep26

Peak month: 36.1%. Bars are scaled to that peak, not to 100%.

Which agents

GitHub CopilotGitHub39727.8%
Claude CodeAnthropic241.7%

Verify it yourself

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

  • aad07ef2Add FP8/FP4 block-quantized support to GatherBlockQuantized contrib op (#3GitHub Copilot
  • 84d758f8Suppress preload errors for optional CUDA libraries (#32112)GitHub Copilot
  • e8787e28Skip compiling Flash Attention .cu sources when onnxruntime_USE_FLASH_ATTEGitHub Copilot
  • 44100ea4Give the CUDA plugin build the CUDA 13 CCCL include path (#32392)Claude Code
  • 09f2e562Force fp32 fallback for CPU-assigned fp16 nodes with no matching kernel (#Claude Code
  • 15c25549Fix GCC 15 -Werror build warnings (#29679)Claude Code

What was excluded

Commits in window1,479
Merge commits removed0
CI / dependency bots removed51
Considered1,428

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 microsoft/onnxruntime

Scan another repository