Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →小a助手配置模板。基于 xiaomo-starter-kit 改编，提供预配置的 OpenClaw 助手框架文件。当用户需要快速配置新助手、设置助手身份、创建助手配置文件时使用此技能。
.claude/skills/majiayu000-xiaomo-assistant-template/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -29% | 0% |
基于 xiaomo-starter-kit 改编的 OpenClaw 助手配置模板，帮助老板aa快速配置个性化的 AI 助手。
根据用户需求生成完整的助手配置文件套件：
快速开始：
bash# 复制模板到新目录 cp -r skills/xiaomo-assistant-template/templates/* /path/to/new-assistant/
或让小a助手帮你生成：
| 文件 | 用途 | |------|------| | SOUL.md | 定义助手的核心性格和行为边界 | | USER.md | 记录用户信息，让助手更了解你 | | IDENTITY.md | 设定助手的名字、形象和说话风格 | | HEARTBEAT.md | 配置助手定期自动执行的检查项 | | MEMORY.md | 助手的长期记忆存储 |
以下 skills 可从 ClawdHub 安装，与本模板配合使用：
bash# 天气查询 clawdhub install weather # 自然语言提醒 clawdhub install remind-me # 任务管理 clawdhub install jdrhyne/todo-tracker # Google 套件（需 OAuth） clawdhub install gog # YouTube 视频摘要 clawdhub install youtube-watcher
原模板来自 xiaomo-starter-kit，由孟健创建。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-18 | fail→pass | 14,900 | 2,087 | -86% | 1 | 1 | 0% | 1,608 | 877 | -45% | 0 | 0 | — |
case-01 | fail→pass | 27,670 | 19,172 | -31% | 1 | 1 | 0% | 3,392 | 2,756 | -19% | 0 | 0 | — |
case-02 | fail→pass | 24,538 | 17,052 | -31% | 1 | 1 | 0% | 2,867 | 3,179 | +11% | 0 | 0 | — |
case-03 | fail→pass | 24,142 | 18,818 | -22% | 1 | 1 | 0% | 3,034 | 2,785 | -8% | 0 | 0 | — |
case-04 | fail→pass | 13,524 | 3,105 | -77% | 1 | 1 | 0% | 1,405 | 993 | -29% | 0 | 0 | — |
case-05 | fail→pass | 14,119 | 16,300 | +15% | 1 | 1 | 0% | 2,240 | 1,488 | -34% | 0 | 0 | — |
case-06 | pass→pass | 14,543 | 2,088 | -86% | 1 | 1 | 0% | 1,436 | 822 | -43% | 0 | 0 | — |
case-07 | fail→pass | 13,445 | 7,376 | -45% | 1 | 1 | 0% | 1,631 | 922 | -43% | 0 | 0 | — |
case-08 | fail→pass | 15,910 | 2,918 | -82% | 1 | 1 | 0% | 1,645 | 808 | -51% | 0 | 0 | — |
case-09 | fail→pass | 16,976 | 3,314 | -80% | 1 | 1 | 0% | 1,879 | 962 | -49% | 0 | 0 | — |
case-10 | fail→pass | 15,056 | 8,128 | -46% | 1 | 1 | 0% | 1,474 | 820 | -44% | 0 | 0 | — |
case-11 | fail→pass | 17,690 | 7,780 | -56% | 1 | 1 | 0% | 1,835 | 887 | -52% | 0 | 0 | — |
case-12 | fail→pass | 13,220 | 2,830 | -79% | 1 | 1 | 0% | 1,966 | 996 | -49% | 0 | 0 | — |
case-13 | fail→pass | 8,001 | 7,471 | -7% | 1 | 1 | 0% | 1,290 | 832 | -36% | 0 | 0 | — |
case-14 | fail→pass | 17,997 | 2,775 | -85% | 1 | 1 | 0% | 2,016 | 914 | -55% | 0 | 0 | — |
case-15 | fail→pass | 13,047 | 4,087 | -69% | 1 | 1 | 0% | 1,752 | 895 | -49% | 0 | 0 | — |
case-16 | fail→pass | 11,095 | 1,750 | -84% | 1 | 1 | 0% | 915 | 705 | -23% | 0 | 0 | — |
case-17 | fail→pass | 8,165 | 2,361 | -71% | 1 | 1 | 0% | 445 | 872 | +96% | 0 | 0 | — |
case-19 | fail→pass | 15,457 | 16,719 | +8% | 1 | 1 | 0% | 2,602 | 2,419 | -7% | 0 | 0 | — |
case-20 | pass→fail | 11,474 | 15,249 | +33% | 1 | 1 | 0% | 1,854 | 2,214 | +19% | 0 | 0 | — |
case-21 | pass→fail | 9,139 | 13,382 | +46% | 1 | 1 | 0% | 1,832 | 2,111 | +15% | 0 | 0 | — |
case-22 | pass→pass | 12,376 | 7,821 | -37% | 1 | 1 | 0% | 1,942 | 1,879 | -3% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +73 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.