ClawHub
Back to skill

Security audit

ComfyUI绘画免费版

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a local ComfyUI image-generation helper, but its activation language and execution scope are broader than its stated purpose.

Review this skill before installing. Use it only for local ComfyUI text-to-image work, keep output paths explicit, avoid using it for general LLM or agent-orchestration tasks, and prefer binding ComfyUI to localhost unless you intentionally want LAN exposure.

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)

Vague Triggers

High
Confidence
92% confidence
Finding
The activation text is overly broad and overlaps with unrelated AI, agent orchestration, and LLM tasks, which can cause the skill to be invoked outside its intended image-generation scope. In an agent environment with exec/write capabilities, misrouting tasks to this skill increases the chance of unintended command execution, filesystem changes, or use in contexts the user did not authorize.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The capability coverage section uses expansive keyword matching rather than precise activation criteria, making accidental or overbroad invocation more likely. In a tool-using agent, ambiguous triggers can route unrelated requests into a skill that has local execution and file-writing side effects.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill states that images are automatically saved locally but does not prominently warn users about filesystem side effects or obtain confirmation for output destinations. In an agent setting with write capability, silent file creation can lead to unexpected data placement, overwrites, storage abuse, or persistence of sensitive/generated content.

Static analysis

No suspicious patterns detected.