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.

slime is an LLM training framework for RL Scaling.

01Result

THUDM/slime ↗

THUDM/slime AI coding statistics: 16%

of 375 commits in February–September 15, 2026 were committed by an AI agent — a floor, not an estimate

16% agent · top contributor 68.3%14 contributors

team

A small group is using agents. Adoption has spread beyond one person but is not project-wide.

agent commits

315

other considered commits

agent contributors

68.3%

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

60assisted

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

Peak month: 52.6%. Bars are scaled to that peak, not to 100%.

Which agents

Agent	Vendor	Commits	Share
GitHub Copilot	GitHub	43	11.5%
Claude Code	Anthropic	14	3.7%
Cursor Agent	Anysphere	3	0.8%

03Evidence

Inspect the matches

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

Matched commit samples

05370bf9 on GitHubFix PYTHONBUFFERED typo to PYTHONUNBUFFERED=1 (#1967)GitHub Copilot
2a89c169 on GitHub[WIP] fix loss oom (#1788)GitHub Copilot
5d7ca976 on GitHub[release] bump to v0.2.4 (#1777)GitHub Copilot
aed93211 on GitHubfix(mtp): support multi-head MTP loss logging (mtp-num-layers > 1) (#2132)Claude Code
19abf05b on GitHubdocs(examples): list coding_agent_rl in examples/README (#2133)Claude Code
21b1b33d on GitHubdocs: drop dangling Dr.GRPO custom-reducer example reference (#2096)Claude Code
16c15fc2 on GitHubfix: reject non-positive rollout temperature at parse time (#2297)Cursor Agent
1da1bb19 on GitHubfix(opd): score teacher logprobs at rollout temperature, not 0 (#2085)Cursor Agent
a0d6d26a on GitHubfix(train): skip optimizer and scheduler for eval-only (#2296)Cursor Agent

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 window	375
Merge commits removed	−0
CI / dependency bots removed	−0
Considered	375

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 →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 THUDM/slime

Scan another repository

Try

Month	Agent share	Considered commits
2026-02	8.5%	47
2026-03	52.6%	78
2026-04	8.8%	34
2026-05	5.4%	74
2026-06	5.1%	59
2026-07	0%	26
2026-08	9.1%	55
2026-09	0%	2