Please do not open a public issue for security problems.
Use GitHub's private vulnerability reporting:
https://github.com/lingmao233/wechat-agent-linux/security/advisories/new
Include the affected version, a minimal reproduction, and the impact. We will acknowledge the report and coordinate a fix before public disclosure.
This project automates a real WeChat client. Use a dedicated test account, not
your primary account. Real sends are gated by WECHAT_ALLOW_SEND; platform
writes default to dry-run.
The repository ignores .env, databases, logs, browser sessions, screenshots,
and local state. If a credential is exposed, rotate the VM password, model API
keys, and the test WeChat session; then remove the secret from Git history
before publishing a fork.
Only the latest release on main is supported with security fixes.