62.5%

of 1,958 commits were committed by an AI agent — a floor, not an estimate

62.5% agent · top contributor 19.2%54 contributors

org-wide

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

1,224

agent commits

734

human commits

agent contributors

19.2%

largest contributor

Assisted or autonomous

1169assisted

55autonomous

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: 76.8%. Bars are scaled to that peak, not to 100%.

Which agents

Other / generic agent	Various	1119	57.2%
Claude Code	Anthropic	92	4.7%
Cursor Agent	Anysphere	32	1.6%
OpenAI Codex	OpenAI	4	0.2%
Devin	Cognition	4	0.2%
GitHub Copilot	GitHub	2	0.1%

Verify it yourself

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

bab1baf2fix(dev): stop baking absolute VITE_BACKEND_BASE_URL in npm run dev (#1660Other / generic agent
dc99e986fix: preserve backend scope in conversation links (#16091)Other / generic agent
fa21e01afix(onboarding): preselect OpenHands LLM provider after picking OpenHands Other / generic agent
708ce3d1fix(automations): surface non-MCP-installable integrations instead of sileClaude Code
a2fef1d7fix(settings): de-duplicate LLM Profile attributes in the "All" tab (#1615Claude Code
13952d79feat: surface conversation tags in the conversation panel (#1926)Claude Code
98155d1bfeat(chat): add inline markdown artifact previews (#16185)Cursor Agent
73434d68fix: make sidebar "Load more" discover folders, not folder contents (#1617Cursor Agent
b7e325e9fix: make overflow menus usable on touch devices (#16101)Cursor Agent
1f8a848crefactor(types): use client error events (#16214)OpenAI Codex
d4f825c7[codex] Fix OpenAI subscription device code contrast (#1531)OpenAI Codex
74f06866Use libraries for local proxy and static serving (#1543)OpenAI Codex
932edbf8feat(backends): compact Cloud vs Agent-server add-backend chooser (#16211)Devin
9025e810feat(home): group workspace folders by parent in the dropdown (#129) (#136Devin
c95176f5settings: add parallel tool calls (tool_concurrency_limit) to Agent SettinDevin
2e1502f3fix: spawn launcher services without implicit shell (#16093)GitHub Copilot
ef1a3c71fix(app-server): preserve observability context metadata (#15215)GitHub Copilot

What was excluded

Commits in window	2,053
Merge commits removed	−1
CI / dependency bots removed	−94
Considered	1,958

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 All-Hands-AI/OpenHands

Scan another repository

Try