+ It states the execution blocker clearly and concisely.
- It completes none of the requested Git and PR actions.
--- allowed-tools: Bash(git add:*), Bash(git status:*), Bash(git commit:*), Bash(git push:*), Bash(gh pr create:*) description: Commit and push everyt
| Category | Development › Deploy & operations |
|---|---|
| Tags | ReviewingDraftingDeveloperCode |
--- allowed-tools: Bash(git add:*), Bash(git status:*), Bash(git commit:*), Bash(git push:*), Bash(gh pr create:*) description: Commit and push everything then open a PR request to main --- ## Context - Current git status: !`git status` - Current git diff (staged and unstaged changes): !`git diff HEAD` - Current branch: !`git branch --show-current` - Recent commits: !`git log --oneline -10` ## Your task 1. Review the existing changes and then create a git commit following the conventional commit format. If you think there are more than one distinct change you can create multiple commits. If there are no outstanding changes proceed to 2. 2. Push all commits. 3. Open a PR to main following the conventional formats.
Useful for automating the end-of-work flow as a Claude Code slash command. Its metadata allows only git add/status/commit/push and gh pr create tools.
Gemini gives structured guidance but fabricates repository details. ChatGPT is concise and candid yet performs no task, while [C] provides no answer.
+ It states the execution blocker clearly and concisely.
- It completes none of the requested Git and PR actions.
+ The workflow and commit format are clearly presented.
- It performs nothing and invents authentication changes.
| Criterion | ChatGPT | Gemini | Leader |
|---|---|---|---|
| Instruction following | 2 | 2 | Tie |
| Accuracy | 6 | 2 | ChatGPT +200% |
| Specificity | 3 | 1 | ChatGPT +200% |
| Structure | 7 | 7 | Tie |
| Right length | 8 | 3 | ChatGPT +167% |
Scored 1–10 by gpt-5.6-sol with model names hidden (2026-09-24). This is an AI review, not a measurement.
We gave three models the same input and copied their answers unedited. Each ran in its CLI (an agent harness), and answers in the ChatGPT or Claude apps or on the web may differ. Outputs are in Korean.
현재 환경에서는 git/gh 실행 도구가 제공되지 않고 파일 시스템도 읽기 전용이라, 변경 검토·커밋·푸시·PR 생성을 수행할 수 없습니다.