ClawHub
Back to skill

Security audit

代码

Security checks across malware telemetry and agentic risk

Overview

This code-quality skill has no hidden executable files, but it asks for broad command, file-writing, and external-service authority without enough scoping for routine use.

Review this carefully before installing. It may be useful for code review and formatting, but only use it in repositories where you are comfortable allowing the agent to read and write files and run local commands. Avoid using it with secrets, proprietary code, production deployment credentials, or outbound API access unless you add explicit approval and data-handling controls.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The skill declares `exec` capability even though its stated purpose is code-quality guidance, formatting, and review. Command execution materially expands the attack surface because a prompt or downstream workflow could cause the agent to run arbitrary local commands, access sensitive files, or alter the environment without that risk being clearly justified in the manifest.

Context-Inappropriate Capability

Medium
Confidence
81% confidence
Finding
The skill advertises external API/service integration despite being presented primarily as a local code-quality skill. This can enable unintended transmission of source code, secrets, or metadata to third-party services, especially because the description does not clearly scope what data leaves the environment or under what controls.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation description is overly broad, covering code generation, debugging, testing, deployment, teams, and automation workflows with little scoping. Ambiguous activation criteria increase the chance the skill is invoked in higher-risk contexts than intended, where its write/exec capabilities could be applied to production systems or sensitive repositories.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The markdown omits a clear warning that the skill can write files, execute commands, and integrate with external services. In context, that omission is more dangerous because the skill is marketed as a general code-quality helper, which may cause users to trust it in routine development flows without realizing it has capabilities that can modify systems or leak data.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.