Skip to main content
skill CX Codex

notify-wecom — Claude Code skill: one-off WeCom messages

Sends a single one-off message to a WeCom (Enterprise WeChat) group bot. Handy when the team needs a quick ping without templates or setup.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add daymade/claude-code-skills --skill notify-wecom -a claude-code
Open on GitHub

What notify-wecom does

The skill pulls the webhook from the shared config at ~/.config/setup-notifications-via-wecom/config.json and posts one message. Plain text only — no markdown cards, images or @mentions. Message length caps at 4096 bytes in UTF-8.

Before sending, it reads recipient_scope and recipient_label. A scope of self sends without an authorization prompt. A scope of others shows the exact label and message first, then waits for human confirmation. A missing or invalid scope is a configuration error: the skill stops and asks you to classify the target instead of guessing.

It is a lightweight companion to setup-notifications-via-wecom. You can install it alone, but then you create the config file by hand. The skill inherits the host and project proxy policy and never forces a direct route. Authorization is not delivery proof — a zero errcode only means the API accepted the call.

What it helps with

  • fire a one-off notification into a WeCom group
  • tell the team a backup or build finished
  • verify the webhook is configured and responding
  • send a multi-line message in one command
  • separate self-delivery from confirmed delivery to others

Example prompts for the agent

  • /notify-wecom Claude 备份完成 ✅
  • /notify-wecom 告警:build 失败,请检查
  • /notify-wecom "line one\nline two\nline three"

How to install notify-wecom

1Skills CLI (Vercel)

npx skills add daymade/claude-code-skills --skill notify-wecom -a claude-code

2Manually: put the files in place

~/.claude/skills/notify-wecom/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/notify-wecom/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about notify-wecom

How do I install notify-wecom in Claude?

Run npx skills add daymade/claude-code-skills --skill notify-wecom -a claude-code. The skill file then lands at ~/.claude/skills/notify-wecom/SKILL.md.

What does notify-wecom do?

It reads the webhook and recipient settings from the shared config and sends one text message to a WeCom group bot. Templating and structured cards are out of scope.

What does the skill need to work?

The file ~/.config/setup-notifications-via-wecom/config.json with webhook_url, recipient_scope and recipient_label. Without it, the skill prints the setup command and stops.

Source text

SKILL.md · MIT
---
name: notify-wecom
description: >-
  Send a single one-off message to a WeCom (Enterprise WeChat) group bot. Use this skill whenever
  the user says "/notify-wecom", "send a quick WeCom message", "企微通知一下", "临时发一条企业微信",
  or any one-shot notification that does not need a reusable template or setup workflow.
argument-hint: "[message]"
---

# /notify-wecom

Send a single message to a WeCom group bot.

## Human confirmation boundary

Read the explicit `recipient_scope` and `recipient_label` from the shared config.
`self` is the user's own delivery channel and may send without authorization.
`others` requires the active human-confirmation gate with the exact label and
message. Missing/invalid scope is a configuration error: stop and classify the
target; do not guess. A local alert/outbox item still needs identity/content
validation before either route.

## Usage

```
/notify-wecom Claude Code 备份完成 ✅
```

The skill reads the webhook URL from the shared config file:

```
~/.config/setup-notifications-via-wecom/config.json
```

If the config is missing, it prints the one-line setup command and stops.

## Prerequisite

This skill is a lightweight companion to `setup-notifications-via-wecom`. Either:

- Install `setup-notifications-via-wecom` first (recommended — it provides the webhook setup steps and its bundled sender), or
- Create the config file manually as shown below.

## What It Does

1. Reads `~/.config/setup-notifications-via-wecom/config.json` for `webhook_url`, `recipient_scope`, and `recipient_label`.
2. Inherits the host/project HTTP(S) proxy and no_proxy policy; never forces direct access.
3. If scope is `self`, sends without an authorization prompt. If scope is `others`, presents the exact label and message for human confirmation.
4. Sends via the WeCom webhook using the sender from `setup-notifications-via-wecom` (or an equivalent inline curl call).
5. Reports success or the exact WeCom error; authorization alone is not delivery proof.

## Examples

```
/notify-wecom 服务上线成功 🚀
/notify-wecom 告警:build 失败,请检查
/notify-wecom 今日同步完成,0 缺失 0 滞后
```

## Multi-line Messages

For multi-line messages, wrap in triple quotes or use a file:

```
/notify-wecom "第一行
第二行
第三行"
```

## Failure Modes

- **Config missing**: Run the setup step from `setup-notifications-via-wecom`, or create the config file manually.
- **Recipient or sender identity missing/invalid**: Use `setup-notifications-via-wecom`'s `set_recipient.py` command; never infer self versus others from the webhook URL, and never trust a script merely because its basename is `send_wecom.py`.
- **Webhook key invalid**: WeCom returns an `errcode`; the skill prints it.
- **Network failure**: Inspect the approved route; do not clear required proxies or switch paths automatically. Follow the sender Skill’s network guidance.

## Limitations

- Plain text only. No markdown cards, images, or @mentions.
- Message length limit is 4096 bytes (UTF-8).
- No templating — use `setup-notifications-via-wecom` for structured backup/alert/status messages.
- A prepared pending item is not a delivery receipt; a zero `errcode` proves API acceptance. Recipient receipt and native SLS delivery require their own evidence.

© daymade · MIT · daymade/claude-code-skills

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set64

llm-eval-harness This skill probes a third-party LLM endpoint across six dimensions: availability, request fidelity, speed, concurrency, protocol compliance and quality regression. CX Codex +1 372installs github-sensitive-data-cleanup Scans and removes secrets, tokens, and personal data from Git history, then verifies and pushes safely. Use it after accidentally leaking credentials to a GitHub repository. CX Codex +1 353installs claude-switch-models-setup Sets up isolated profiles to run multiple providers like Kimi, DeepSeek, and Anthropic in separate terminal windows. It stops configuration drift while sharing common skills and hooks. CX Codex +1 344installs design-style-picker Generates a batch of visual design directions and lets you choose from images instead of a written brief. Handy when a client cannot put the look into words. CX Codex +1 342installs setup-notifications-via-wecom Sets up a WeCom group bot webhook and sends technical status messages through it. Useful when a backup, sync or cron job should report on its own. CX Codex +1 341installs local-codex Runs the local OpenAI Codex CLI as a non-interactive sub-agent using ChatGPT OAuth. Ideal for offloading code generation and reviews without API keys. CX Codex +1 333installs wps-doc-scraper This skill archives public WPS/KDocs documents and embedded ProcessOn mind maps as raw data, SVG or PNG, and Markdown. Useful when a link has to be saved without logging in. CX Codex +1 334installs codex-image-gallery Launches a local browser gallery for reviewing Codex-generated images. Useful when you need to inspect local output folders without leaving your agent workflow. CX Codex +1 326installs

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor