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: 5.3%. Bars are scaled to that peak, not to 100%.
Which agents
| Claude Code | Anthropic | 731 | 2.5% |
| Cursor Agent | Anysphere | 67 | 0.2% |
| GitHub Copilot | GitHub | 17 | 0.1% |
| OpenAI Codex | OpenAI | 14 | 0% |
| Gemini CLI | 1 | 0% |
Verify it yourself
Sample commits that matched. Every one links to GitHub — check the trailers.
b6912258clang: Fix exception model flag tests if webassembly isn't build (#223770)Claude Codea3d2f4b4[[mlir][linalg] Infer reduction-neutral padding values in rewriteAsPaddedOClaude Code88cab6bfRuntimeLibcallsEmitter: Let a consumer's own library variant beat its exclClaude Code90f36e05[OpenMP] Remove the LocationDescription constructor that dropped the locatCursor Agent092ac4c2[clang][OpenMP] Don't drop debug location when handing off to OMPIRBuilderCursor Agentf276d8f0[OMPIRBuilder] Don't leak a foreign debug loc into the taskwait call (#222Cursor Agentdc948be7[runtimes] Escape list separators when forwarding search paths (#216169)GitHub Copilotee1eab2a[clang] Update diagnostics to include matrices as accepted types (#201237)GitHub Copilotbc97c56e[DirectX] Expand {u,s}mul.with.overflow in DXILIntrinsicExpansion (#207297GitHub Copilotb8c7806b[Flang][OpenMP] Fix use_device_addr descriptor types (#222961)OpenAI Codexd450ba4b[AMDGPU] Improve handling of numRecords for make.buffer.rsrc (#213180)OpenAI Codexfe1bfb32[SPIR-V] Lower nested aggregate insertvalue operands (#204239)OpenAI Codex4b5f5954[clang][docs] Add basic user documentation for Hardware-assisted AddressSaGemini CLI
What was excluded
| Commits in window | 29,820 |
| Merge commits removed | −0 |
| CI / dependency bots removed | −181 |
| Considered | 29,639 |
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 llvm/llvm-project