feishu-cli-mail
飞书邮箱专用入口,覆盖收件箱分诊、邮件和线程读取、发送、回复、转发、草稿、查询签名、 CID 内联图片和模板。用户提到飞书邮件、邮箱、收件箱、未读邮件、草稿、邮箱签名、 邮件模板、回复、转发、发送预览/确认或发送 HTML/CID 邮件时必须使用本 Skill。 仅要求预览或等待发送确认也属于本 Skill 的草稿/发送工作流,必须先加载本 Skill 准备预览。 只读命令可用 User/Bot;写入、签名和模板需要 User。聊天消息使用 feishu-cli-messaging。
How do I install this agent skill?
npx skills add https://github.com/riba2534/feishu-cli --skill feishu-cli-mailIs this agent skill safe to install?
- Gen Agent Trust Hubpass
This skill facilitates Feishu email management through a command-line interface. It includes internal safety checks such as mandatory confirmation for sending emails and protection against header injection. The primary security consideration is the inherent risk of processing untrusted content from email bodies which could lead to indirect prompt injection.
- Socketwarn
1 alert: gptSecurity
- Snykwarn
Risk: MEDIUM · 1 issue
- ZeroLeakspass
1 finding · Score: 86/100
What does this agent skill do?
飞书邮箱
读取 references/workflows/mail/workflow.md 后执行。
将该工作流中的 references/ 相对路径按 workflow.md 所在目录解析。
执行规则
triage/message/messages/thread支持--as bot|user|auto;Bot 必须显式指定邮箱,不能用me。写类、签名和模板管理需 User Token。auth check只预检本地 User Token,不作为 Bot 的前置条件。- 用户只要求草稿或尚未明确发送时保存草稿;已明确授权发送且收件人、主题、正文齐备时直接使用
--confirm-send,不重复索取确认。普通附件暂不支持,不要承诺发送未支持的附件。 - 回复和转发前先读取原邮件,避免选错 message ID 或 thread ID。
- 不在日志或结果中回显邮件正文里的敏感信息。
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/riba2534/feishu-cli/feishu-cli-mail">View feishu-cli-mail on skillZs</a>