菜单
Compare Claude Code and Codex on the same real code-change task with isolated worktrees, identical gates, transcripts, time, and cost.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/reidemeister94/development-skills --skill ai-agent-bench命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Use when repository docs, README, AGENTS.md, rules, plans, chronicles, or ATLAS are stale or current work changed their lifecycle; archives task history, with --clean for corpus-wide cleanup.
Research current best practices and turn the evidence into a clear recommendation. Use for state-of-the-art, evidence-based, optimal-approach, comparison, or pros-and-cons questions in any field.
Use when the user wants to do brainstorming, choose an approach, design something or clarify an ambiguous or consequential change before planning or implementation.
基于 SOC 职业分类
正在显示 SKILL.md
| name | ai-agent-bench |
| description | Compare Claude Code and Codex on the same real code-change task with isolated worktrees, identical gates, transcripts, time, and cost. |
| user-invocable | true |
| allowed-tools | Glob, Grep, Read, Bash, Edit, Write |
Compare agents only with the same task, starting commit, and outcome check. The harness preserves result branches and removes temporary worktrees.
Create <repo>/.agent-bench.toml:
prompt = "prompts/task.md"
start_branch = "main" # or start_commit
agents = ["claude", "codex"]
outer_check = "./scripts/full_check.sh"
inner_check = "pytest tests/integration/test_x.py -q"
outer_check proves the real outcome before and after, and measures wall time. inner_check gives agents fast feedback.
Require a clean repo, available CLIs, and a passing outer_check. Confirm agents and run ID, then run trials sequentially to avoid load-biased timing:
python <skill>/scripts/run_trial.py --repo "$REPO" --config "$REPO/.agent-bench.toml" --agent "$AGENT" --run "$RUN_ID"
Results go to eval-results/<task>/<agent>/run-<id>-<timestamp>/. Record unexpected behavior in ai-agent-bench-anomalies.md per anomalies.
Aggregate with scripts/parse_transcript.py --aggregate <run-dirs> --output comparison.json --render-report comparison.md. Report gates, branches, time delta, tokens, and cost. Never rank a failed trial.
For plugin behavior rather than a real code task, use the bounded Pydantic runner documented by eval-regression and scripts/run_evals.py.
Never commit on the user's branch. A repeated run creates a new timestamped result and preserves prior evidence.