Waiting Tips

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:16 PM UTC
• Version:
  1.0.0

Overview

This skill displays local bilingual waiting tips and does not show hidden data access, exfiltration, or destructive behavior.

Install this if you want tips to appear while OpenClaw is responding. Expect additional tip messages in normal chats, choose zh-only or en-only if bilingual output is unwanted, and only run the platform bot examples with accounts you control while keeping bot tokens and WhatsApp auth files private.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Vague Triggers

Medium

• Confidence: 84% confidence
• Finding: The README states that tips appear automatically during normal chat flow and that OpenClaw can call the `waiting_tip` tool before processing requests. Because the activation scope is described broadly and without clear gating conditions, users and integrators may enable a behavior that fires in many conversations, creating unintended tool invocation, extra message noise, or surprise disclosure of plugin behavior across platforms.

Vague Triggers

Medium

• Confidence: 79% confidence
• Finding: The manual examples use very generic phrases like '来个 tip' and 'show me 3 tips', which overlap with ordinary conversational language. In systems that map natural language loosely to tools, vague trigger phrasing can cause accidental invocation when a user is speaking casually rather than intentionally calling the skill.

Vague Triggers

High

• Confidence: 97% confidence
• Finding: Triggering automatically before every AI response is overly broad and causes the skill to run during routine conversations without clear user consent or contextual bounds. This can create persistent unsolicited output, interfere with normal assistant behavior, and become a vehicle for spammy or policy-bypassing message injection across all conversations.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: Manual triggers like 'show me a tip' and 'give me a tip' are common conversational phrases that can easily collide with normal user requests, causing accidental activation. That broad matching increases the chance of unintended responses, user confusion, and prompt-routing mistakes, especially in multi-skill environments.

Natural-Language Policy Violations

Medium

• Confidence: 90% confidence
• Finding: Emitting bilingual Chinese/English output on every wait event without explicit user language preference or opt-in can expose users to unwanted content and degrade usability across channels. While not directly a code-execution issue, it is a genuine security/quality concern because unsolicited cross-language output broadens data presentation and may violate user expectations or platform norms.

VirusTotal

64/64 vendors flagged this skill as clean.