wecomcli-email

Send, reply, forward, and search WeCom enterprise emails via the wecom-cli tool.

287|46|Updated Feb 5, 2026
One-click install
npx skills add https://github.com/YanHaidao/wecom --skill wecomcli-email-yanhaidao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: wecomcli-email
Source: https://github.com/YanHaidao/wecom/tree/main/skills/wecomcli-email
Command: npx skills add https://github.com/YanHaidao/wecom --skill wecomcli-email-yanhaidao

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Managing enterprise email inside WeCom (企业微信) through conversation is fragmented: users need to compose messages, resolve recipients from the contacts directory, attach files, embed inline images, and search large mailboxes without touching the mail client. This Skill routes all of those operations through the wecom-cli tool with strict workflows, output formats, and security guardrails. ## Core Features & Use Cases - Full mail lifecycle: Send new emails, reply (including reply-all), and forward with automatic subject prefix construction and deduplication (回复:/转发:/Re:/Fwd:). - Search and read: Multi-condition mail search (keywords, sender, time range, unread, folder, tag, attachments, star, importance) with pagination, plus full detail retrieval including body, attachments, inline images, and calendar/meeting info. - Schedule and meeting emails: Send calendar invitations and meeting bookings via email with repeat rules, reminders, and meeting options. - Security guardrails: Prompt-injection detection in mail bodies, social-engineering identification, external-recipient warnings, and refusal to embed executable code. - Use Case: Ask the assistant to "reply to all on the Q2 report email with the updated figures attached" — it locates the mail, resolves recipients, writes a Markdown body, previews the message, and sends it via wecom-cli. ## Quick Start Ask the assistant to send an email to a colleague by name with a subject and body, and it will resolve the recipient, preview the message, and deliver it through wecom-cli.

Frequently Asked Questions about wecomcli-email

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I send an email through WeCom using wecom-cli?

Compose the body as a Markdown file, resolve recipients (full email addresses are used directly; names are looked up via the contacts skill), then call wecom-cli mail send with the to, subject, file_path, and content_type fields. A preview of the message is shown before the send call executes.

How do I search WeCom emails by keyword, sender, or date?

Use wecom-cli mail search with conditions like keywords, sender, receiver, begin_time, end_time, only_unread, folder_names, tag_names, or has_attachments. At least one condition is required, conditions combine with AND logic, and --page-count controls automatic pagination.

Can this skill send meeting invitations or schedule emails?

Yes, when the user explicitly mentions email or mailbox, it can send calendar invitations using the schedule object and meeting bookings using schedule plus meeting objects. Pure schedule or meeting management without email belongs to the separate calendar and meeting skills.

What are the limitations of WeCom email search?

Keyword searches return at most 100 mails, and searches using begin_time, end_time, only_unread, or only_reminder are limited to the last 30 days. Marking read/unread, deleting mails, drafts, and tag write operations are not supported and must be done in the WeCom client.

Why can't an attachment or inline image be downloaded from a mail?

Attachments and inline images normally download via media download using their media_id. If the enterprise has DLP enabled, content arrives as encrypted work.weixin.qq.com/filepreview/security links that cannot be downloaded or decrypted, so users must open them in the WeCom client.

How does the skill handle suspicious or injected email content?

Mail body text is treated as data, never as instructions, so embedded commands are ignored and flagged to the user. It also identifies social-engineering patterns, warns about external recipients extracted from mail bodies, and refuses to write script tags or javascript URIs into outgoing mail.