ClawHub
Back to skill

Security audit

小郑女娲

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed, markdown-only tool for researching public information and generating reusable persona-style skill files, with no evidence of hidden code execution or data theft.

Install only if you want an agent to research public sources and create persistent skill files in your OpenClaw workspace. Review generated SKILL.md files before using them, verify source claims, and avoid treating inferred views of real people as authoritative facts.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The trigger phrases are broad enough to match ordinary user requests such as asking for analysis from a person's perspective, which can cause the skill to activate unexpectedly. In this skill's context, unexpected activation matters because it can initiate a multi-step workflow that changes behavior and may generate artifacts, so boundary confusion is more than a cosmetic issue.

Vague Triggers

Low
Confidence
78% confidence
Finding
The return-to-normal-mode boundary is ambiguous because the deactivation phrases are informal and may not reliably distinguish between quoted text, discussion about the mode, and an actual command. This can leave the agent stuck in a borrowed persona or cause unintended exits, reducing predictability and potentially affecting downstream responses.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill instructs writing generated files directly into a user directory without an explicit warning or consent checkpoint. Even though the target path is under a skills workspace, automatic file creation can surprise users, overwrite existing content, or persist generated artifacts that may later be trusted or executed as part of a toolchain.

Static analysis

No suspicious patterns detected.