ClawHub
Back to skill

Security audit

hello-openclaw

Security checks for vulnerabilities and agentic risk

Overview

This is a simple demo greeting skill with overly generic triggers, but its disclosed behavior is low impact and purpose-aligned.

This skill appears safe to install as a toy/demo skill. Users should be aware that its triggers are generic and may fire when they only meant to say hello or run an unrelated test; a publisher should rename the triggers to more specific phrases for clarity.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (4)

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
94% confidence
Finding
The trigger "test openclaw" begins with the built-in command word "test," creating a shadow-command condition where the skill may collide with platform-reserved behavior. This can lead to ambiguous routing, unexpected execution precedence, or user confusion about whether the built-in command or the skill will run.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
97% confidence
Finding
The trigger "run test" directly overlaps with the built-in command word "run" and is highly likely to conflict with native command parsing. A user attempting a normal platform action could unintentionally invoke this skill instead, or the skill could mask expected built-in behavior.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger list includes very generic phrases such as "run test," which can cause the skill to activate in situations unrelated to this greeting skill. Even though the skill behavior is harmless, broad activation increases the chance of unintended invocation and confusion for users or other tooling.

Vague Triggers

Medium
Confidence
97% confidence
Finding
The usage text says the skill can be triggered by mentioning broad terms like "hello," which encourages activation on extremely common language. This makes accidental invocation more likely and may interfere with normal conversation flow or with other skills that use similar phrases.

Static analysis

No suspicious patterns detected.