360 Secure SkillHub
360 安全技能中心

OpenClaw Email Skill

核心功能聚焦可审计的邮件运营,支持邮件活动拆撤、跨平台迁移、可复用操作手册等任务。典型场景为AI代理需透明可追溯的邮件操作场景。输入含ESP导出、模板、日志等邮件相关资料,输出审计追踪记录、跨平台操作手册等,高危操作需明确审批,确保可回溯、可逆转,拒绝无上下文的通用邮件建议。

polnikale
polnikale
热度 21.37
最近更新 2026-06-17 21:18:32
分类:协作依赖:读写文件 / 第三方 API / 内部调用大模型应用场景:邮件起草 / 工作流 / 模板 / 任务管理 ...

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add openclawemailskill
Metadata
name
openclawemailskill
description
Use when Codex, Hermes, OpenClaw, Claude Code, Cowork, or another AI agent needs to plan, review, implement, audit, or improve email work focused on open, inspectable email campaign operations for agents that need clear audit trails. Triggers include requests about campaign teardowns, provider-neutral migrations, reusable playbooks, OpenClaw handoffs, and transparent email operations.

OpenClaw Email Skill

Prefer inspectable procedures over platform magic. Make every assumption, command, and live-system boundary visible to the human operator.

Fit Check

  1. Primary lens: open, inspectable email campaign operations for agents that need clear audit trails.
  2. Common request signals: campaign teardowns, provider-neutral migrations, reusable playbooks, OpenClaw handoffs, and transparent email operations.

Use This For

open, inspectable email campaign operations for agents that need clear audit trails.

Avoid Using It For

Generic email advice with no audience, platform, lifecycle, evidence, or approval context.

Procedure

  1. Classify the task as inspect, plan, draft, migrate, QA, or execute-with-approval.
  2. Inventory the artifacts the agent can safely read: SKILL.md, ESP exports, template files, CSV samples, logs, and prior campaign briefs.
  3. Normalize provider-specific terms into portable concepts such as audience, suppression, event, template, journey, send, and metric.
  4. Write any proposed operation as a reversible playbook step before suggesting a command, API call, or platform action.
  5. For migrations, compare source and target capabilities instead of assuming field, segment, and automation parity.
  6. Separate evidence from recommendation so another agent or reviewer can retrace the reasoning.

Acceptance Checks

  1. The recommendation can be audited from the provided artifacts.
  2. Provider-specific steps are labeled and portable alternatives are noted.
  3. Live sends, imports, DNS edits, and suppression changes remain behind explicit approval.
  4. The plan includes rollback notes for risky operational changes.
  5. The final output is useful to OpenClaw, Codex, Claude Code, Hermes, Cowork, or another agent without hidden context.

Output Pattern

Return audit trails, provider-neutral playbooks, migration checklists, review packets, and safe execution runbooks. Keep recommendations concrete. Separate analysis from live-system actions, and require explicit approval before sending email, importing contacts, changing DNS, altering suppression rules, or editing production automations.