Chinese Name Generator - Cantian AI
三才五格姓名分析与起名推荐技能。用于用户请求“分析这个名字的三才五格”“算姓名五格吉凶”“结合喜用神推荐名字”“按姓氏起单字/双字名”等场景;关键词包括:三才五格、姓名分析、起名、喜用神、天格、人格、地格、外格、总格、吉凶。 / Three-talents and five-grid Chinese naming...
MIT-0 · Free to use, modify, and redistribute. No attribution required.
⭐ 2 · 32 · 0 current installs · 0 all-time installs
bycantian.ai@tianlinle
MIT-0
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name, description, SKILL.md and the included scripts/data all implement San Cai & Wu Ge analysis and name candidate generation. Required files (data/hanzi.json, analyzeName.ts, pickName.ts, wuge logic) match the stated purpose; nothing requested or present appears unrelated (no cloud creds, no unexpected binaries).
Instruction Scope
Runtime instructions are limited to running the included TypeScript/Node scripts against the local hanzi.json dataset. Scripts validate inputs and explicitly error on unsupported characters. They do not reference or read system paths outside the skill directory, environment variables, or external endpoints.
Install Mechanism
There is no install spec embedded in the skill (instruction-only). SKILL.md suggests running `npm i` and optionally installing `tsx` for compatibility; package.json contains no dependencies so `npm i` is effectively a no-op unless the user adds packages. Installing `tsx` would pull from the public npm registry (normal for Node tools) but is optional and not required by the skill itself.
Credentials
The skill requests no environment variables, credentials, or config paths. The code does not access process.env or other secrets; all data used is the bundled hanzi.json and internal logic.
Persistence & Privilege
Skill flags are default (always: false). The skill does not request permanent platform presence or modify other skill/system configurations. It runs as a local CLI tool when invoked.
Assessment
This skill appears coherent and local-only: it reads the bundled hanzi.json and runs numeric rules to analyze and score Chinese names. Before running, inspect the data file if you have concerns about included characters. You don't need any credentials. If you follow the SKILL.md advice to run `npm i` and optionally `npm i -D tsx`, be aware those commands will contact the npm registry to fetch packages (the skill itself has no declared dependencies). Run the tool in an isolated environment or review the code if you plan to allow it to run automatically in an agent. If you want additional assurance, run the scripts manually in a sandboxed Node environment and verify there are no unexpected network calls (there are none in the provided source).Like a lobster shell, security has layers — review code before you run it.
Current versionv0.0.2
Download ziplatest
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
SKILL.md
三才五格姓名分析 / Name Analysis (San Cai & Wu Ge)
- Brand:
Cantian AI - Primary Site: https://cantian.ai
何时使用 / When to Use
- 用户要分析中文姓名的三才五格(天格/人格/地格/外格/总格)。
- 用户要判断姓名格数吉凶,或查看每格对应五行。
- 用户要验证单字名和双字名是否都能计算(如
李明、欧阳若曦)。
前置依赖 / Prerequisites
- 推荐运行环境:Node 24(可直接运行 TypeScript 源码)
- 兼容方案:若 Node 版本较低,使用
tsx执行 - 执行目录:在 skill 根目录(
SKILL.md所在目录)运行以下命令 - 脚本按 TypeScript 源码直接运行,不需要预编译
npm i
# 仅在需要兼容运行时安装
npm i -D tsx
脚本清单 / Script Index
scripts/analyzeName.ts:分析指定姓名的三才五格结果scripts/pickName.ts:按姓氏与喜用神五行筛选并打分推荐候选名字(单字/双字)
脚本与参数 / Scripts & Parameters
scripts/analyzeName.ts
# 推荐方式
node scripts/analyzeName.ts --surname <姓> --given <名>
# 兼容方式(fallback)
tsx scripts/analyzeName.ts --surname <姓> --given <名>
参数定义:
--surname(必填):中文姓氏;长度 1-2 个字符;无默认值;缺失、超长或为空时报错并退出--given(必填):中文名字(不含姓氏);长度 1-2 个字符;无默认值;缺失、超长或为空时报错并退出--help(选填):打印使用说明后退出- 不支持未知参数;传入未知参数时报错并退出
输出说明:
- 默认输出 Markdown 报告,包含:
- 基础信息(姓、名、全名)
- 用字明细(简体/康熙字形/拼音/笔画/汉字五行)
- 五格结果(天格、人格、地格、外格、总格的数值、吉凶、数理五行)
- 三才关系摘要(天-人-地五行组合、天人关系、人地关系)
错误行为:
- 姓名中任一字不在
data/hanzi.json时,脚本报错并退出(非 0) - 参数缺失、参数值非法或存在未知参数时,脚本报错并退出(非 0)
scripts/pickName.ts
# 推荐方式
node scripts/pickName.ts --surname <姓> [--given-len <1|2|both>] [--favorable-element <金|木|水|火|土>] [--secondary-element <金|木|水|火|土>] [--allow-unknown-element] [--allow-level2] [--disable-name-filter]
# 兼容方式(fallback)
tsx scripts/pickName.ts --surname <姓> [--given-len <1|2|both>] [--favorable-element <金|木|水|火|土>] [--secondary-element <金|木|水|火|土>] [--allow-unknown-element] [--allow-level2] [--disable-name-filter]
参数定义:
--surname(必填):中文姓氏;长度 1-2 个字符;缺失或非法时报错并退出--given-len(选填):候选名长度;取值1|2|both;默认both;非法值时报错并退出--favorable-element(选填):喜用神主五行;取值金|木|水|火|土--secondary-element(选填):喜用神次五行;取值金|木|水|火|土--allow-unknown-element(选填):启用后,element缺失的字在五行筛选时可参与候选;默认关闭--allow-level2(选填):启用后可纳入level=2字;默认仅使用level=1常用字--disable-name-filter(选填):关闭“人名友好过滤”;默认开启--help(选填):打印使用说明后退出
输出说明:
- 默认输出 Markdown 报告,包含:
- 输入参数摘要
- 候选池统计(字池大小、单字/双字生成数量、返回数量)
- 候选列表(总分、分数拆解、用字属性、五格结果、三才关系)
- 脚本会返回一批候选名;对用户回复时不暴露具体候选数量
筛选与打分规则:
- 双轨筛选:同时考虑“汉字五行(喜用神)+ 三才五格数理”
- 若指定喜用神五行,默认排除
element缺失的字;启用--allow-unknown-element后可放开 - 默认只使用
level=1常用字并启用“人名友好过滤”;可用--allow-level2、--disable-name-filter放宽 - 吉凶、三才生克关系、字级别(
level)共同参与分数计算 - 双字名结果会做分散控制,避免候选过度集中在少数字形组合
示例 / Examples
# 单字名最小可用示例
node scripts/analyzeName.ts --surname 李 --given 明
# 双字名最小可用示例
node scripts/analyzeName.ts --surname 欧阳 --given 若曦
# pickName 最小可用示例(单字+双字)
node scripts/pickName.ts --surname 李
# pickName 指定喜用神
node scripts/pickName.ts \
--surname 李 \
--given-len both \
--favorable-element 木 \
--secondary-element 火
推荐执行流程 / Recommended Flow
- 先运行
pickName.ts获取一批候选。 - 让大模型从候选中筛出“好听、顺口、语义正向”的少量名字(例如 3-8 个),并说明筛选理由。
- 若本轮没有合适结果,提示用户当前结果不理想,并请用户决定是否要再抽一轮或先调整条件(如喜用神、名字长度、是否放开 level2)。
注意事项 / Notes
- 所有命令必须在本 skill 根目录执行,不依赖仓库根目录路径。
analyzeName.ts只做姓名分析;候选推荐请使用pickName.ts。- 文中“五行”分两类:
数理五行(来自三才五格数值映射)与汉字五行(来自字库字段element),两者不可混用。 - 五格计算基于
wugeStrokeCount;若字形存在异体字争议,结果以字库记录为准。 - 单姓/单名按传统规则补 1:单姓天格补 1,单名地格补 1。
- 面向用户的文案不要写“共筛出 N 个候选”这类具体数量,也不要列举被剔除的字(如负面语义字);仅呈现最终推荐名及推荐理由。
Files
9 totalSelect a file
Select a file to preview.
Comments
Loading comments…
