Thanks for your interest in contributing! / 感谢你的贡献意愿!
git clone https://github.com/xyva-yuangui/smartness-eval.git
cd smartness-eval
# Verify skill structure / 验证技能结构
python3 scripts/check.py
# Run a quick eval to confirm everything works / 确认评估可运行
python3 scripts/eval.py --mode quick --no-probesRequirements / 环境要求:
- Python 3.9+
- No external packages needed (stdlib only) / 无需额外依赖
| Area | Examples | 示例 |
|---|---|---|
| Scoring formulas | Better metric blending, new evidence signals | 更好的指标融合方式 |
| Test definitions | New tests in config/task-suite.json |
新测试用例 |
| Safety rules | Stricter command validation patterns | 更严格的安全校验 |
| Visualization | HTML report, chart generation | 报告可视化 |
| Documentation | Translations, tutorials, examples | 翻译、教程、示例 |
| Bug fixes | Edge cases, fallback handling | 边界情况修复 |
One PR = one logical change. Avoid mixing unrelated changes.
- No unsafe command execution patterns introduced / 未引入不安全的命令执行
- No absolute-path assumptions / 未使用绝对路径
- No
exec(),eval(), or-cpatterns / 无动态代码执行 -
python3 scripts/check.pypasses / 健康检查通过 -
python3 scripts/eval.py --mode quick --no-probesruns successfully / 快速评估可运行 -
CHANGELOG.mdupdated if user-facing change / 用户可感知的变更需更新 CHANGELOG - Documentation updated if needed / 如需要则更新文档
- New test commands only use
python3interpreter - All file paths are relative and within allowed prefixes
- No new external network calls unless explicitly opt-in
- No hardcoded credentials or API keys
Use Conventional Commits:
feat: add HTML report renderer
fix: handle missing cron-governor-report gracefully
docs: add Chinese translation for FAQ
test: add edge case for empty reasoning store
chore: update CI workflow to Python 3.12
security: block additional unsafe subprocess patterns
Please include / 请包含以下信息:
| Field | Description | 说明 |
|---|---|---|
| OpenClaw version | e.g., 2026.3.13 | OpenClaw 版本 |
| OS | e.g., macOS 15.3, Ubuntu 24.04 | 操作系统 |
| Python version | e.g., 3.11.4 | Python 版本 |
| Exact command | Full command used | 完整命令 |
| Error output | Full traceback / stderr | 完整错误输出 |
| Data context | Which state files exist/missing | 状态文件情况 |
- Follow existing patterns in
eval.py - Type hints for function signatures
- Docstrings for public functions
- No external dependencies (stdlib only)
Open an issue or start a GitHub Discussion. We're happy to help!
有问题请提 Issue 或在 GitHub Discussions 讨论,我们很乐意帮助!