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.

Evaluate and improve models and agents using environments

01Result

NVIDIA-NeMo/Gym ↗

NVIDIA-NeMo/Gym AI coding statistics: 17.7%

of 1,039 commits in February–September 15, 2026 were committed by an AI agent — a floor, not an estimate
17.7% agent · top contributor 19%50 contributors
org-wide

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

184
agent commits
855
other considered commits
50
agent contributors
19%
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

184assisted
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
2026-022.9%68
2026-036.2%97
2026-0414.5%83
2026-0527.3%110
2026-0621%143
2026-0724.3%136
2026-0820.6%281
2026-0910.7%121

Peak month: 27.3%. Bars are scaled to that peak, not to 100%.

Which agents

AgentVendorCommitsShare
Claude CodeAnthropic17416.7%
Cursor AgentAnysphere101%
OpenAI CodexOpenAI50.5%

03Evidence

Inspect the matches

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

Matched commit samples

  • 5b34fcda on GitHubfeat(birdbench): BM25-driven schema context with PK/FK info, plus SQL-commClaude Code
  • d109dec6 on GitHubfix(verifiers_agent): never store the policy server's session cookie (#331Claude Code
  • fddaf235 on GitHubfix(nemotron_3.5_super): make P/D router topology and policy configurable Claude Code
  • d6d7d976 on GitHubfeat: add inference_provider server for hosted inference providers (#1286)Cursor Agent
  • 9aefd55c on GitHubdocs: Link example server READMEs to environment tutorials (#1011)Cursor Agent
  • 520e47f8 on GitHubdocs: improve getting started (#1283)Cursor Agent
  • a47fc3e0 on GitHubfeat(sandbox): add Docker Compose adapter (#3332)OpenAI Codex
  • 5cc2e761 on GitHubfix(sandbox): silence OpenSandbox SDK info logs (#2550)OpenAI Codex
  • a5944740 on GitHubAdd workplace-claude demo notebook and config (#2101)OpenAI Codex

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 window1,039
Merge commits removed−0
CI / dependency bots removed−0
Considered1,039

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 →Cursor Agent commitsBrowse repositories with detected Cursor Agent signatures →OpenAI Codex commitsBrowse repositories with detected OpenAI Codex signatures →Python AI coding statisticsCompare evidence across Python 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 NVIDIA-NeMo/Gym

Scan another repository

Try