Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Universal Occupation Adapter

v1.0.0

通用职业适配器 —— 输入任何职业名称,自动生成完整的职业专用认知Skill,让SOUL哲学覆盖所有职业

0· 85·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for kingofzhao/universal-occupation-adapter.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Universal Occupation Adapter" (kingofzhao/universal-occupation-adapter) from ClawHub.
Skill page: https://clawhub.ai/kingofzhao/universal-occupation-adapter
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install universal-occupation-adapter

ClawHub CLI

Package manager switcher

npx clawhub@latest install universal-occupation-adapter
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The SKILL.md describes a Python class API (UniversalOccupationAdapter), install commands, auto-publish and batch-generate behaviors, and generation of new Skill packages — but the published bundle contains no code files or install spec. That mismatch (promised programmatic module + install but no code) is incoherent: either required code is missing or the documentation misrepresents what the skill actually provides.
!
Instruction Scope
Runtime instructions explicitly instruct web searches (Wikipedia/GitHub), file generation (SKILL.md/VERIFICATION_PROTOCOL/HEARTBEAT/README), self-verification, batch generation, and optional auto-publish to ClawHub. Those actions involve network access and creating/publishing artifacts. The instructions do not request credentials or explicitly limit destinations, so an agent following them could contact external sites and attempt to publish results without clear safeguards.
Install Mechanism
There is no install spec in the bundle (instruction-only), which is lowest technical risk. However the README/SKILL.md advertise 'clawhub install' and a cp-based manual install; because no install spec or packaged code exists, those commands are misleading and would fail or have unclear effects.
!
Credentials
The skill declares no required environment variables or credentials, yet its examples and features (auto_publish to ClawHub, GitHub/Wikipedia searches, potential API usage for verification/publishing) imply needing service credentials or tokens. The absence of declared credentials vs. implied external integrations is a mismatch that could lead to unexpected credential prompts or failures.
Persistence & Privilege
The skill does not request always-on presence and has defaults that allow user invocation and autonomous invocation. HEARTBEAT.md describes periodic checks for queued requests, implying background polling, but the bundle provides no mechanism to install such a daemon. No explicit system-wide privileges or modifications to other skills are requested.
What to consider before installing
This skill's documentation promises a Python API, automatic generation, verification, and publishing of new 'occupation' skills, but the package contains only instructions and templates — no executable code or declared credentials. Before installing or invoking it: - Ask the author for the missing implementation (the UniversalOccupationAdapter module) or a clear disclaimer that the SKILL.md is a high-level spec only. - Do not enable auto_publish/auto_verify or any batch auto-run flags until you know where outputs will be sent and whether publishing requires credentials. - Expect the adapter to perform web searches and create files in the workspace; run it in a sandboxed environment or review generated files before publishing. - If you intend to use publishing features (ClawHub/GitHub), require explicit documentation about what credentials are needed and how they are stored; prefer personal tokens with least privilege. - If you need high assurance, request unit tests or an actual implementation package and review that code before granting network access or tokens.

Like a lobster shell, security has layers — review code before you run it.

latestvk971zsf4t6xvr3gm4e2s5kw13583yv7a
85downloads
0stars
1versions
Updated 4w ago
v1.0.0
MIT-0

Universal Occupation Adapter

元数据

字段
名称universal-occupation-adapter
版本1.0.0
作者KingOfZhao
发布日期2026-03-31
置信度96%

来源碰撞

skill-collision-engine (碰撞引擎)
        ⊗
programmer-cognition (程序员适配示例)
        ⊗
researcher-cognition (科研适配示例)
        ⊗
ai-growth-engine (成长引擎)
        ↓
universal-occupation-adapter (通用适配器)

核心哲学

SOUL 五律是通用认知框架,不依赖任何特定领域。 每个职业的差异不在框架本身,在 5 个维度的填充内容。 本适配器输入职业名,输出完整的职业专用 Skill。

职业五维度模型

每个职业都可以用 5 个维度完整描述:

维度1: known_sources    — 这个职业的"已知"从哪里来?
维度2: unknown_types     — 这个职业的"未知"是什么类型?
维度3: verification_methods — 这个职业如何验证正确性?
维度4: memory_types      — 这个职业需要什么样的"文件记忆"?
维度5: redlines          — 这个职业的红线是什么?

SOUL 五律映射:
  1. 已知/未知    ← 维度1 + 维度2
  2. 四向碰撞     ← 维度3(验证方法的不同视角)
  3. 人机闭环     ← 维度3(人类实践验证)
  4. 文件即记忆   ← 维度4
  5. 置信度+红线  ← 维度5

已验证的职业模板

从 programmer-cognition 和 researcher-cognition 中提取的模板:

occupation_templates = {
    "程序员": {
        "known_sources": ["API文档", "代码库", "依赖关系", "运行环境"],
        "unknown_types": ["运行时行为", "并发安全", "边缘case", "第三方服务行为"],
        "verification_methods": ["单元测试", "CI Pipeline", "Code Review", "生产监控"],
        "memory_types": ["docstring", "CHANGELOG", "debug日志", "架构文档"],
        "redlines": ["不硬编码密钥", "不裸except", "不跳过测试", "不操作生产DB",
                     "不删除数据(trash>rm)", "不在周五部署"]
    },
    "科研人员": {
        "known_sources": ["已发表论文", "实验数据", "已验证理论", "可复现结果"],
        "unknown_types": ["未验证假设", "矛盾数据", "理论空白", "方法局限"],
        "verification_methods": ["统计显著性", "同行评审", "可复现性检查", "对照组实验"],
        "memory_types": ["literature_review/", "hypotheses.md", "experiments/", "insights/"],
        "redlines": ["不伪造数据", "不cherry-pick", "不忽略矛盾数据",
                     "不复制不引用", "不发布未验证结论"]
    },
    "设计师": {
        "known_sources": ["设计系统", "品牌规范", "用户画像", "竞品分析"],
        "unknown_types": ["用户真实感受", "跨设备一致性", "可访问性", "文化差异"],
        "verification_methods": ["设计系统检查", "A/B测试", "用户测试", "可访问性审计"],
        "memory_types": ["design_log/", "iteration_history/", "component_library/", "user_research/"],
        "redlines": ["不忽视可访问性", "不忽略用户反馈", "不抄袭设计",
                     "不跳过移动端检查", "不使用未授权字体/素材"]
    },
    "企业家": {
        "known_sources": ["市场数据", "财务报表", "用户反馈", "竞争对手动态"],
        "unknown_types": ["市场真实需求", "时机判断", "团队执行力", "外部宏观变化"],
        "verification_methods": ["MVP验证", "市场反馈", "财务指标", "用户留存数据"],
        "memory_types": ["decision_log/", "pivot_history/", "market_analysis/", "financial_model/"],
        "redlines": ["不烧钱盲目扩张", "不忽视现金流", "不欺骗用户/投资者",
                     "不忽略竞争信号", "不在数据不足时做重大决策"]
    },
    "教师": {
        "known_sources": ["课程标准", "学生基础数据", "教学经验", "学科知识"],
        "unknown_types": ["学生真实理解程度", "最有效的教学方式", "个体差异需求", "注意力状态"],
        "verification_methods": ["随堂测验", "作业分析", "期末评估", "学生反馈"],
        "memory_types": ["teaching_log/", "student_progress/", "lesson_plans/", "assessment_data/"],
        "redlines": ["不放弃任何一个学生", "不体罚/言语侮辱", "不照本宣科",
                     "不延迟反馈", "不泄露学生隐私"]
    },
    "医生": {
        "known_sources": ["临床指南", "患者病史", "检查结果", "医学文献"],
        "unknown_types": ["个体差异反应", "罕见病例", "药物相互作用", "患者依从性"],
        "verification_methods": ["随访结果", "同行会诊", "临床指南对照", "患者反馈"],
        "memory_types": ["case_log/", "differential_diagnosis/", "treatment_protocols/", "continuing_education/"],
        "redlines": ["不误诊(二次确认)", "不过度治疗", "不忽视患者主诉",
                     "不泄露患者隐私", "不超范围执业"]
    }
}

适配算法

def adapt_to_occupation(occupation_name: str, custom_dimensions=None) -> Skill:
    """
    输入: 职业名称 + 可选自定义维度
    输出: 完整的职业专用 Skill
    
    算法:
    1. 查找已有模板 → 如果找到,直接使用
    2. 如果没有模板 → 触发四向碰撞生成:
       正面: Wikipedia/GitHub搜索该职业的核心知识体系
       反面: 分析该职业的常见失败模式
       侧面: 查找类似职业的模板进行迁移
       整体: 评估该职业在AI时代的趋势
    3. 将5个维度填充到SOUL五律框架中
    4. 生成SKILL.md + VERIFICATION_PROTOCOL + HEARTBEAT + README
    5. 自验证(置信度≥95%才输出)
    """

支持无限扩展

已有模板: 程序员, 科研人员, 设计师, 企业家, 教师, 医生 (6个)
自动生成: 任何职业 → 四向碰撞 → 新模板 → 验证 → 发布

示例自动生成请求:
  "律师" → 正面(法规体系) + 反面(常见败诉原因) + 侧面(类似医生模板迁移) + 整体(AI法律趋势)
  "建筑师" → 正面(建筑规范) + 反面(常见结构失误) + 侧面(类似设计师模板迁移) + 整体(AI辅助设计趋势)
  "产品经理" → 正面(需求分析方法) + 反面(常见产品失败模式) + 侧面(类似企业家模板迁移) + 整体(AI产品趋势)

安装命令

clawhub install universal-occupation-adapter
# 或手动安装
cp -r skills/universal-occupation-adapter ~/.openclaw/skills/

调用方式

from skills.universal_occupation_adapter import UniversalOccupationAdapter

adapter = UniversalOccupationAdapter(workspace=".")

# 1. 使用预设模板
skill = adapter.generate(
    occupation="程序员",
    output_dir="./skills/programmer-cognition"
)

# 2. 自动生成新职业Skill(无预设模板时触发四向碰撞)
skill = adapter.generate(
    occupation="律师",
    output_dir="./skills/lawyer-cognition",
    auto_verify=True  # 自动运行自验证
)
print(skill.confidence)  # 0.96
print(skill.new_insights) # ["法庭辩论可用四向碰撞", ...]

# 3. 自定义维度
skill = adapter.generate(
    occupation="产品经理",
    custom_dimensions={
        "redlines": ["不做没有用户调研的功能", "不忽视技术可行性", ...]
    }
)

# 4. 列出所有可用模板
templates = adapter.list_templates()
# ["程序员", "科研人员", "设计师", "企业家", "教师", "医生"]

# 5. 批量生成(一键生成N个职业Skill)
results = adapter.batch_generate(
    occupations=["律师", "建筑师", "产品经理", "心理咨询师"],
    auto_publish=True  # 自动验证+发布到ClawHub
)

与其他 Skill 的关系

SOUL (根)
  └── universal-occupation-adapter (本Skill: 职业适配器)
        ├── programmer-cognition (已生成)
        ├── researcher-cognition (已生成)
        ├── [设计师版] (可用本Skill生成)
        ├── [企业家版] (可用本Skill生成)
        ├── [教师版] (可用本Skill生成)
        └── [任何职业] (四向碰撞自动生成)

学术参考文献

  1. A Survey of Self-Evolving Agents — 自进化框架通用性
  2. SAGE: Multi-Agent Self-Evolution — 多领域适应
  3. Group-Evolving Agents — 经验迁移(←跨职业模板迁移)
  4. Self-evolving Embodied AI — 领域自适应
  5. Memory in the Age of AI Agents — 职业记忆差异化
  6. Beyond RAG for Agent Memory — 跨领域知识聚合

Comments

Loading comments...