21%

of 334 commits were committed by an AI agent — a floor, not an estimate

21% agent · top contributor 11.4%44 contributors

org-wide

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

agent commits

264

human commits

agent contributors

11.4%

largest contributor

Assisted or autonomous

65assisted

5autonomous

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

Peak month: 36.4%. Bars are scaled to that peak, not to 100%.

Which agents

Claude Code	Anthropic	59	17.7%
Devin	Cognition	5	1.5%
Cursor Agent	Anysphere	3	0.9%
OpenAI Codex	OpenAI	2	0.6%
GitHub Copilot	GitHub	1	0.3%

Verify it yourself

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

2028eeb0docs: reference PR number in changelog entryClaude Code
cccbcf4bfix: price the Claude 5 model family for local cost estimatesClaude Code
d4152844feat: add display.rightAlign for merged expanded linesClaude Code
f262936dtest(cost): clarify Bedrock test intent and add real-session coverage (#63Devin
dd3f4471Add unit tests for modules with lowest coverage (#635)Devin
a549cfebImprove error handling: add debug logging to silent catch blocks (#633)Devin
939eb664docs(contributing): write down the HUD scope bar (#738)Cursor Agent
901c1bf5fix: agents, git locks, cache wording, and zh docs (#737)Cursor Agent
25c2f1b8feat(cost): add opt-in daily cumulative cost (#736)Cursor Agent
c6e7e665release: 0.4.0 (#660)OpenAI Codex
f5c4ea28Fix symlinked rules and compact cache tokens (#659)OpenAI Codex
e8d64924fix: resilient usage display under API rate limiting (#193)GitHub Copilot

What was excluded

Commits in window	574
Merge commits removed	−93
CI / dependency bots removed	−147
Considered	334

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 jarrodwatts/claude-hud

Scan another repository

Try