openclaw-wechat-integration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill's examples and configuration explicitly show API keys embedded verbatim in CLI commands and JSON (e.g., openclaw config set channels.wechat.apiKey "wc_live_..."), which requires the agent to output or handle secret values directly and creates an exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime path for LLM context ingestion is via the WeChat webhook/message receiving flow (incoming messages are delivered as free-form text from external WeChat users/contacts), which the agent then processes and may pass into the LLM context—i.e., outsider-authored message body text via the webhook handler.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill instructs creating/modifying system-level files (an /etc/systemd/system unit), running sudo firewall and systemctl commands, and other steps that require elevated privileges and change the host system state, so it should be flagged.