emailautomationskill Skill
当 Codex、Hermes、OpenClaw、Claude Code、Cowork 或其他 AI 代理需要规划、评审、实现、审计或改进侧重行为触发器、生命周期旅程、自动化治理和运营保障的邮件工作时使用。触发场景包括触发器清单、自动化 QA、旅程优化、过期流程清理、事件驱动的生命周期修复和治理评审。
安装方式:把技能目录放入 ~/.claude/skills/(Claude Code)或在 claude.ai 设置中启用;也可复制右侧安装命令一键添加。
技能指令原文(SKILL.md)
Email Automation Skill
Treat automation as production infrastructure. Triggers, exits, exclusions, and approval gates matter as much as copy.
Agent Contract
The agent may inspect, reason, draft, critique, and prepare handoffs for behavioral triggers, lifecycle journeys, automation governance, and operational safeguards.
- Primary lens: behavioral triggers, lifecycle journeys, automation governance, and operational safeguards.
- Common request signals: trigger inventory, automation QA, journey optimization, stale-flow cleanup, event-driven lifecycle fixes, and governance reviews.
- Default posture: Treat automation as production infrastructure. Triggers, exits, exclusions, and approval gates matter as much as copy.
- Useful output family: automation contracts, trigger inventories, QA plans, overlap audits, governance checklists, and journey optimization recommendations.
- Production boundary: separate recommendation from execution.
- Evidence boundary: say which source material supports the recommendation.
- Review boundary: identify the human owner for risky changes.
- Data boundary: do not assume missing fields, consent, or suppression state.
- Platform boundary: describe provider-specific steps in operational language.
- Measurement boundary: define what success or recovery will look like.
- Handoff boundary: make the next step clear to another agent.
Use Cases
Typical requests include trigger inventory, automation QA, journey optimization, stale-flow cleanup, event-driven lifecycle fixes, and governance reviews.
Runbook
- Identify whether the automation is behavioral, lifecycle, transactional, reactivation, operational, or sales-assisted.
- Write the journey contract: trigger, eligibility, first action, waits, branches, exits, suppression, and success event.
- Check data freshness and event reliability before recommending new branches or personalization.
- Audit overlapping automations for duplicate sends, contradictory offers, and frequency-cap pressure.
- Propose governance rules: naming, ownership, change log, test contacts, rollback, and scheduled review cadence.
- Require explicit approval for activation, deactivation, imports, suppression edits, and any production journey change.
Quality Bar
- The trigger reflects behavior, not vague intent.
- Exits remove contacts when the journey is no longer relevant.
- Fallbacks exist for missing data and delayed events.
- Automation changes include test cases and rollback notes.
- Governance keeps future agents from editing a live journey blindly.
Output Pattern
Return automation contracts, trigger inventories, QA plans, overlap audits, governance checklists, and journey optimization recommendations. 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.