How to Integrate OpenClaw with Enterprise WeChat Bot for Automatic Message Broadcasting

Background

OpenClaw is an open‑source AI personal‑assistant framework that can execute desktop tasks (e.g., flight search, copywriting, video creation) via natural‑language commands. The article demonstrates automating message broadcasting to multiple Enterprise WeChat groups.

Prerequisites

API key for a cloud LLM (MiniMax M2.5 or Kimi K2.5)

Cloud server or local computer

Enterprise WeChat administrator account

Model selection

MiniMax M2.5 and Kimi K2.5 provide the best results for tool invocation. A local model was tested and was three minutes per simple command and often failed to control desktop tools; therefore a cloud model is recommended for typical hardware.

Obtain a MiniMax key from the MiniMax Open Platform and note the secret.

Enterprise WeChat bot configuration

In the Enterprise WeChat client: Workplace → Smart Robot → Create Robot → API mode . Choose “Long Connection”, then record the generated Bot ID and Secret.

OpenClaw installation

One‑click script

curl -fsSL https://openclaw.ai/install.sh | bash

iwr -useb https://openclaw.ai/install.ps1 | iex

Run the script and wait for completion.

Manual Node.js installation

Download Node.js ≥ 22, install, then switch npm to the domestic mirror:

npm config set registry https://registry.npmmirror.com

Install OpenClaw globally: npm install -g openclaw@latest Installation usually finishes within a minute.

Configuration wizard

Execute: openclaw onboard --install-daemon Select MiniMax (or Moonshot AI for Kimi) and paste the API key. Skip other options and open the Web UI.

WeChat channel setup

Install the WeChat plugin:

openclaw plugins install @wecom/wecom-openclaw-plugin

Restart OpenClaw and add a channel:

openclaw gateway restart

openclaw channels add

Select “Enterprise WeChat”, enter Bot ID and Secret, choose “Pairing”, and paste the pairing key displayed in the WeChat conversation to complete binding.

Real‑world tests

Mobile interaction

After starting the OpenClaw service, messages can be sent and received from the WeChat mobile client, confirming bot operation on both PC and mobile.

Group broadcast

Two WeChat groups are created. Sending a single message to OpenClaw forwards the same content to both groups. Using the exact chatid (obtained by @‑mentioning the robot or checking OpenClaw logs) yields reliable delivery, whereas using the group name may fail.

Cost considerations

Each command triggers a call to the underlying LLM, incurring token usage. The author observed a single message costing about ¥6; complex workflows can quickly exceed ¥100. Vague prompts increase tool invocations and token consumption, further raising expenses.

Reference

Enterprise WeChat supports OpenClaw integration.