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.

An open-source Harness Engineering platform for coding agents—define harnesses as code, run controlled experiments, inspect evidence, and compare outcomes. Turn task evidence into actionable team and organization insights.

01Result

QoderAI/better-harness ↗

QoderAI/better-harness AI coding statistics: 43.6%

of 500 commits in February–September 15, 2026 were committed by an AI agent — a floor, not an estimate

43.6% agent · top contributor 52.3%22 contributors

team

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

218

agent commits

282

other considered commits

agent contributors

52.3%

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

218assisted

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: 48.6%. Bars are scaled to that peak, not to 100%.

Which agents

Agent	Vendor	Commits	Share
OpenAI Codex	OpenAI	179	35.8%
Claude Code	Anthropic	38	7.6%
GitHub Copilot	GitHub	1	0.2%

03Evidence

Inspect the matches

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

Matched commit samples

c8f20fcf on GitHubfeat(studio): add actionable memory recap skillOpenAI Codex
5fe7136a on GitHubfeat(desktop): add macOS Go esbuild XPC linkerOpenAI Codex
011ce0fa on GitHubfix(studio): keep performance source readable when zoomedOpenAI Codex
6a711e0c on GitHubfix(studio): make a boxed Agent's paths and PATH actually workClaude Code
eb7344b1 on GitHubfeat(studio): show what an Agent is doing before it can answerClaude Code
c382a0f5 on GitHubdocs(studio): mark the one-boxed-run-at-a-time limit as supersededClaude Code
f25cb466 on GitHubfix(pi): support Oh My Pi (OMP) session format and directory naming (#143)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 window	580
Merge commits removed	−67
CI / dependency bots removed	−13
Considered	500

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 →OpenAI Codex commitsBrowse repositories with detected OpenAI Codex signatures →JavaScript AI coding statisticsCompare evidence across JavaScript 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 QoderAI/better-harness

Scan another repository

Try

Month	Agent share	Considered commits
2026-07	39.6%	48
2026-08	48.6%	288
2026-09	36%	164