ClawHub

Wx Huitu

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed chart-generation workflow that creates PNG chart files and can sync them to Feishu/Lark cloud storage.

Install this if you want a workflow that turns article/data content into PNG charts. Treat generated charts as potentially sensitive: use local-only handling or skip Feishu/Lark sync when working with unpublished, internal, or confidential data, and confirm where lark-cli is authenticated before uploading.

SkillSpector

By NVIDIA
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 (5)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The workflow includes automatic Feishu/Lark cloud-drive synchronization even though the skill’s stated purpose is chart generation and PNG delivery. Uploading generated files off-device can exfiltrate article content, embedded data, or sensitive business material without an explicit user-consent gate, which is especially risky for unpublished or internal documents.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The trigger list includes broad, everyday phrases such as “画图” and “做个图”, which can match many ordinary user requests beyond the intended chart-generation scope. This can cause unintended activation of a skill that performs file generation and downstream delivery actions, increasing the chance of accidental data processing or invocation in the wrong context.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill explicitly instructs saving outputs to the desktop and syncing them to Feishu cloud, but does not require a clear user warning or opt-in for these data-affecting actions. Because the input may be article content or data descriptions that could be sensitive, automatic local persistence and cloud upload can lead to unintended disclosure, retention, or policy violations.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The trigger phrases include very broad everyday terms such as “画图” and “做个图,” which can cause the skill to activate in contexts unrelated to chart generation. Over-broad activation increases the chance that unrelated user content is parsed and processed, potentially leading to unintended file generation, data handling, or downstream upload behavior.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The workflow describes cloud-drive synchronization but does not warn the user that generated files will be uploaded off-device. This undermines informed consent and can expose sensitive chart contents, article text, or derived business insights to external storage without the user realizing it.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal