ClawHub
Back to skill

Security audit

Testcase Generator

Security checks across malware telemetry and agentic risk

Overview

This skill is a local test-case generator/updater whose file reads and writes fit its stated purpose, but users should be careful with its in-place library update feature.

Install only if you want a local workflow that reads configured testcase libraries and saves PRD-derived analysis/output files under ~/.openclaw. Before using the update command, verify the configured workbook path, keep a backup of the source library, and review any update workbook manually.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill clearly instructs the agent to read local configuration and case-library files and to write summaries and Excel outputs, yet it declares no permissions. This creates a trust and containment gap: users and the platform may not realize the skill can access and modify local data, increasing the chance of unintended file exposure or writes.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The advertised behavior focuses on analyzing PRDs and generating two Excel outputs, but the workflow also includes reading product configs, scanning local Excel libraries, maintaining metadata, and even overwriting existing case libraries by UUID. That mismatch can cause users to authorize what looks like simple document generation while the skill performs broader local file access and modification than expected.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill writes multiple artifacts to fixed local workspace and memory paths, but the user-facing description does not clearly warn about this persistence. Hidden local writes can expose sensitive PRD content, test plans, and internal product data to other tools or users sharing the same environment, and can leave residual artifacts unexpectedly.

Missing User Warnings

Medium
Confidence
79% confidence
Finding
The script saves modifications directly back to the configured workbook path without any confirmation, backup, atomic write strategy, or integrity check. In the context of a testcase-generation skill, this makes accidental or unintended destructive data modification more dangerous because user-supplied update files can permanently alter the canonical testcase library.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.