ClawHub
Back to skill

Security audit

Task Prism

Security checks for vulnerabilities and agentic risk

Overview

This is a project-planning prompt skill with no executable code or sensitive system access, though its README includes a remote image and its activation wording is broad.

Reasonable to install if you want a structured project-planning assistant. Be aware that viewing the README may load an external group-chat image, and the skill may activate on broad requests like task breakdowns or project plans.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (3)

Context-Inappropriate Capability

Low
Confidence
93% confidence
Finding
The README loads an external image from a third-party domain, which causes network access when the file is rendered and leaks viewer metadata such as IP address, user agent, and access timing. In an agent/skill ecosystem this is unnecessary to the skill’s task-decomposition purpose and creates a supply-chain/tracking surface if the remote content changes or is used for telemetry.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The trigger phrases are broad enough to match ordinary planning requests, which can cause the skill to activate in contexts the user did not explicitly intend. Over-broad invocation increases the chance of prompt hijacking of general conversations, unnecessary context capture, and misrouting to a specialized skill that may shape outputs inappropriately.

Vague Triggers

High
Confidence
92% confidence
Finding
The skill declares extremely broad trigger phrases such as generic requests for task breakdown or project planning, which can cause it to activate on many ordinary user prompts outside its intended scope. Overbroad activation increases the chance of inappropriate routing, response hijacking of unrelated conversations, and reduced reliability of higher-priority or safer skills.

Static analysis

No suspicious patterns detected.