Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed Markdown guide for creating OpenClaw plugins and hooks, with no hidden code, remote execution, credential handling, or destructive behavior found.
Install only if you intend to author OpenClaw plugins that can affect prompt construction. Review any hook code carefully before enabling it, because before_prompt_build hooks can change system context or user messages across future OpenClaw runs.
The trigger phrase 'modify gateway' is ambiguous and could be interpreted as authorization to alter routing, middleware, or request-processing behavior without clear boundaries. In this skill's context, that ambiguity is more dangerous because it is explicitly about hooks that can inject context and modify messages before prompt construction.
The trigger phrase 'modify gateway' is ambiguous and could be interpreted as authorization to alter routing, middleware, or request-processing behavior without clear boundaries. In this skill's context, that ambiguity is more dangerous because it is explicitly about hooks that can inject context and modify messages before prompt construction.
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
mkdir -p ~/.openclaw/plugins/<plugin-name>/hooks
The file presents the skill's natural-language description and operating instructions exclusively in Chinese, which can amount to forcing a specific language without user opt-in. There is no indication that this skill is region-specific or that users may choose another language.
No suspicious patterns detected.