instances/xiaodazi/skills/style-learner/SKILL.md
Learn and memorize user writing style from conversation history and sample texts. Applies learned style to future writing tasks.
npx skillsauth add malue-ai/dazee-small style-learnerInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
从用户的对话和文本样本中学习写作风格,并在后续写作任务中复用。
当用户提供文本样本时,提取风格特征:
输入:用户的 1-3 篇文本样本
↓
分析维度:
- 语气:正式/半正式/口语/学术
- 句式:长句偏好/短句偏好/混合
- 用词:书面/通俗/专业术语密度
- 结构:段落长度/转折词/列表偏好
- 特殊习惯:口头禅/常用表达/标点习惯
↓
输出:风格画像(结构化 JSON)
将风格画像保存为本地文件,供后续写作时参考。
# 风格画像存储路径
~/.xiaodazi/styles/
# 文件格式
~/.xiaodazi/styles/{用户名}_default.json
~/.xiaodazi/styles/{用户名}_{场景名}.json # 可按场景保存多套风格
风格画像结构:
{
"name": "默认风格",
"created_at": "2025-01-15",
"updated_at": "2025-02-07",
"sample_count": 3,
"profile": {
"tone": "半正式,亲切但不随意",
"sentence_style": "短句为主,偶尔用长句做总结",
"vocabulary": "通俗用词,避免学术术语,喜欢用比喻",
"structure": "开头直入主题,段落短(3-5句),善用列表",
"special_habits": [
"喜欢用破折号做补充说明",
"段尾常用反问句引发思考",
"数字用阿拉伯数字不用中文"
]
},
"few_shot_examples": [
{
"context": "写产品介绍",
"sample": "(截取的 200 字典型片段)"
}
]
}
写作任务时,读取风格画像并注入到提示词:
用户请求:「帮我写一篇关于 AI 的文章」
↓
读取风格画像 → 提取关键约束
↓
生成时遵循:
- 语气:半正式,亲切
- 句式:短句为主
- 用词:通俗,多用比喻
- 结构:开头直入主题,段落 3-5 句
↓
输出:符合用户风格的文章
每次用户修改输出或给出反馈时,更新风格画像:
mkdir -p ~/.xiaodazi/styles
cat > ~/.xiaodazi/styles/default.json << 'EOF'
{
"name": "默认风格",
...
}
EOF
cat ~/.xiaodazi/styles/default.json
ls ~/.xiaodazi/styles/
development
Local web search (Tavily/Exa, requires API Key). For quick searches. If no Key configured or deep research needed, use cloud_agent instead.
development
Get current weather and forecasts (no API key required).
tools
Send WhatsApp messages to other people or search/sync WhatsApp history via the wacli CLI (not for normal user chats).
tools
Start voice calls via the Moltbot voice-call plugin.