ClawHub
Back to skill

Security audit

2.5D横幅生成工具

Security checks for vulnerabilities and agentic risk

Overview

This image-generation skill is not malicious, but it asks for broad read, write, and command execution authority without tightly scoped runtime instructions.

Review before installing. Use this only in a controlled workspace, with non-sensitive input images and a scoped API key. Be cautious about allowing command execution unless you can verify the actual script or command path it will run.

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

Medium
Confidence
95% confidence
Finding
The manifest grants `read`, `write`, and `exec` to a skill whose stated purpose is image generation/editing, which exceeds what is justified by the documented functionality. In an agent setting, unnecessary command and filesystem capabilities enlarge the attack surface and can be abused to run arbitrary commands, overwrite files, or access sensitive local data if later combined with prompt injection or unsafe command construction.

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The documentation explicitly broadens the skill from image generation into generic file handling, API integration, and command execution, which is inconsistent with the narrow banner/image-editing use case. This encourages use of powerful primitives outside the intended scope, making misuse and privilege abuse more likely in automated agent workflows.

Vague Triggers

Medium
Confidence
81% confidence
Finding
The activation/use description is vague and overly broad, including generic wording about image create/modify requests without clear boundaries or input constraints. Ambiguous activation criteria increase the chance that an agent invokes this skill in inappropriate contexts, potentially exposing powerful tools like filesystem access or command execution when not needed.

Static analysis

No suspicious patterns detected.