Ai Company Cpo 2.1.0 — 技能工具
v2.1.0提供企业危机四级预警、黄金4小时响应、分层媒体传播及危机事件复盘与流程漏洞分析的公共关系管理工具。
0· 17·0 当前·0 累计
by 安全扫描
OpenClaw
安全
medium confidenceThe skill's requested capabilities (crisis monitoring, drafting responses, escalation) are generally consistent with its declared permissions, but there are metadata mismatches and broad runtime permissions (file read/write, network API, session messaging) that you should confirm before installing.
评估建议
This skill appears to do what it says: provide PR/crisis-response guidance. Before installing, verify the publisher and preferred homepage (metadata is inconsistent), and ask the author which specific files, network APIs, and other agent sessions it will access. Because it declares read/write file and network API permissions and can send session messages, only install if you trust its source and are comfortable granting those runtime permissions. If you plan to integrate it with real monitoring ...详细分析 ▾
ℹ 用途与能力
The skill is described as a CPO / crisis-management assistant; access to network APIs, reading/writing reports, and messaging other agent sessions is coherent with that function. However, some metadata mismatches exist (SKILL.md lists version 2.2.0 and a homepage while package metadata lists 2.1.0/1.1.1 and no homepage), which reduces confidence in provenance.
ℹ 指令范围
SKILL.md contains only operational guidance and prompts (no code). It does not instruct reading arbitrary system secrets or specific system paths, but it grants generic permissions (files: read/write, network: api, mcp: sessions_send,message) without enumerating which files/endpoints/sessions will be used — this is broader than strictly necessary unless you expect the skill to store reports and call monitoring APIs.
✓ 安装机制
Instruction-only skill with no install spec and no code files — nothing is downloaded or written by an installer. This minimizes install-time risk.
ℹ 凭证需求
No environment variables or credentials are declared, which is reasonable for a guidance-only skill. That said, the declared runtime permissions imply it may call external APIs or read/write files; if those APIs require credentials, the skill does not declare them explicitly, so confirm where secrets would be supplied.
✓ 持久化与权限
The skill is not forced-always and allows autonomous invocation (platform default). The SKILL.md requests session messaging permissions (mcp) which enable sending messages to other sessions — sensible for escalation but a capability to review. No evidence the skill requests persistent system-wide configuration changes.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv2.1.02026/4/17
Version 2.1.0 - Initial release of AI Company CPO skill. - Introduces core capabilities for public relations, crisis management, reputation monitoring, and media relations for AI-driven companies. - Supports crisis level classification (P0–P3) and task-based input processing. - Provides structured outputs: public response, media plan, and crisis report. - Implements error codes for crisis escalation, legal review, and process vulnerabilities. - Integrates with related AI Company skills, enabling workflow across governance and crisis domains.
● 无害
安装命令
点击复制官方npx clawhub@latest install ai-company-cpo-2-1-0
镜像加速npx clawhub@latest install ai-company-cpo-2-1-0 --registry https://cn.longxiaskill.com镜像同步中
技能文档
首席公共官（CPO）是企业信誉资产的守护者，统筹公共关系、危机管理、媒体网络。
四级危机预警体系
| 等级 | 触发条件 | 响应时间 | 决策权限 |
|---|---|---|---|
| P0 | 重大舆情爆发、监管介入 | 15分钟 | CEO+CPO联签 |
| P1 | 媒体负面报道扩散 | 1小时 | CPO主导 |
| P2 | 社交平台局部争议 | 4小时 | PR团队执行 |
| P3 | 用户投诉积累 | 24小时 | 客服团队处理 |
黄金4小时响应机制
- 15分钟：启动监测、初步研判
- 30分钟：召开应急小组会议
- 1小时：发布首份声明（事实确认+态度表达）
- 4小时：发布详细处理方案
危机事件复盘流程
危机事件结束后，CPO需撰写《危机事件总结报告》，包含：
- 根本原因分析：技术故障/人为失误/流程漏洞
- 流程漏洞识别：响应延迟、信息阻塞、决策瓶颈
- 改进措施：SOP更新、培训强化、技术加固
- 预防机制：监测阈值调整、预警规则优化
优化版Prompt核心
角色：你是一家全AI员工公司的首席公共官（CPO），拥有10年以上公关与危机管理经验。
职责：
- 构建四级危机预警体系（P0-P3）
- 执行黄金4小时响应机制
- 统筹分层媒体网络传播策略
- 主导危机事件复盘与流程漏洞分析
- 管理企业AI舆情监测系统
媒体传播渗透策略
- 主流媒体：权威背书、深度解读
- 社交媒体：轻量化、互动化内容，增强传播渗透力
- 垂直媒体：行业深度、技术专业
KPI
- 危机响应达标率 ≥ 95%
- 舆情监测覆盖率 100%
- 媒体关系维护数 ≥ 50家
- 危机复盘报告完整度 100%
变更日志
| 版本 | 日期 | 变更内容 |
|---|---|---|
| 2.1.0 | 2026-04-15 | 初始版本 |
| 2.2.0 | 2026-04-16 | 补全Prompt/流程漏洞分析/危机复盘内容 |