Pulse
2026-09-18-v3.0
Dataset
2026-10-27-catalog-ai-v2
Ruleset
b2e8889c66f7
Scanned
2026-10-27
Window
February–September 15, 2026
Catalog tracking observation. This repository is measured with the same fixed window and ruleset, but it is not part of the published benchmark or its aggregate claims.

A high-performance inference engine for LLM, VLM, DiT and REC models, optimized for diverse AI accelerators. It is hosted in OpenAtom Foundation.

01Result

xLLM-AI/xllm ↗

xLLM-AI/xllm AI coding statistics: 4.7%

of 991 commits in February–September 15, 2026 were committed by an AI agent — a floor, not an estimate
4.7% agent · top contributor 23.4%17 contributors
org-wide

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

47
agent commits
944
other considered commits
17
agent contributors
23.4%
largest contributor

02Drivers

What shapes this result

Mode, timing, and detected signatures show whether the result reflects broad adoption or one concentrated workflow.

Assisted or autonomous

47assisted
0autonomous

Assisted — a developer’s identity authored the commit and the agent added itself as co-author. Autonomous — a bot identity authored the commit without a human identity attached.

Agent share by month

Agent-attributed commit share by month
MonthAgent shareConsidered commits
2026-020%65
2026-031.9%156
2026-041%99
2026-054%126
2026-065.3%132
2026-0710.7%150
2026-087.1%140
2026-094.1%123

Peak month: 10.7%. Bars are scaled to that peak, not to 100%.

Which agents

AgentVendorCommitsShare
Claude CodeAnthropic242.4%
Gemini CLIGoogle121.2%
Cursor AgentAnysphere111.1%
GitHub CopilotGitHub10.1%

03Evidence

Inspect the matches

Public commit samples let you inspect the signatures behind the aggregate result.

Matched commit samples

  • ad9689e9 on GitHubbugfix: apply the ACL graph batch limit per DP rank in warmup plan.Claude Code
  • 7e928471 on GitHubfeat: add Qwen3-VL-dense to Python model backend (ViT + deepstack + mRoPE)Claude Code
  • 787151da on GitHubrefactor: consolidate platform typing into single torch-based module. (#21Claude Code
  • e13eef88 on GitHubbugfix: fix DeepSeek V4 schedule_overlap + mtp. (#1782)Gemini CLI
  • 9fe5dbde on GitHubfeat: add MUSA layers for Qwen3.5 dense models. (#1885)Gemini CLI
  • ef6b10cd on GitHubfeat: add RL deep-sleep for co-located training. (#1783)Gemini CLI
  • c6b40e06 on GitHubperf: split the api_service ingress benchmarks by hop. (#2330)Cursor Agent
  • cbe14aac on GitHubfeat: add python sfa dcp decode path with naive torch remap.Cursor Agent
  • 478dda20 on GitHubperf: cut copies and lazily allocate logprob buffers in request build pathCursor Agent
  • 050fc251 on GitHubfeat: support dspark for qwen3 on NPU. (#2132)GitHub Copilot

04Method

What entered the denominator

The result begins with commit history, removes merges and routine automation, and reports literal agent signatures as a lower bound.

Exclusion ledger

Commits in window991
Merge commits removed−0
CI / dependency bots removed−0
Considered991

Bots are excluded from the denominator entirely. Agent attribution requires a literal supported signature; no code-style inference is used.

Continue exploring

Claude Code commitsBrowse repositories with detected Claude Code signatures →GitHub Copilot commitsBrowse repositories with detected GitHub Copilot signatures →Cursor Agent commitsBrowse repositories with detected Cursor Agent signatures →Gemini CLI commitsBrowse repositories with detected Gemini CLI signatures →C++ AI coding statisticsCompare evidence across C++ repositories →GitHub AI commit analyzerInspect another repository with the same evidence model →
Reproduce this observation

Run the same detector locally.

Deterministic signature matching; no code-style inference. Squash merges can remove trailers, so this result remains a measured floor.

npx qmmit-cli xLLM-AI/xllm

Scan another repository

Try