ClawHub
Back to skill

Security audit

j-patent-strategy-analyzer

Security checks across malware telemetry and agentic risk

Overview

This skill appears benign: it generates a patent strategy report and clearly discloses its package installation, web/MCP lookups, and checkpoint output files.

Before installing, confirm you are comfortable authorizing the PatSnap/Zhihuiya MCP integration and installing the named Python packages in the runtime. Avoid using highly confidential patent or business files unless the configured platform and MCP account are approved for that data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Context-Inappropriate Capability

Medium
Confidence
91% confidence
Finding
The skill requires the agent to install Python packages into the runtime via `runtime.apply_sync`, expanding its execution environment and introducing supply-chain and environment-modification risk unrelated to a minimally scoped report skill. Although the packages listed are common data-science libraries, automatic dependency installation can be abused or can violate least-privilege expectations if users approve broad runtime changes.

Description-Behavior Mismatch

Medium
Confidence
88% confidence
Finding
The skill advertises a 'zero external dependency' standalone output, but its workflow explicitly depends on web/MCP retrieval and PyPI installation during execution. This mismatch can mislead users and reviewers about the skill's real trust boundary, causing them to approve network access, tool use, or environment changes they would otherwise scrutinize more carefully.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.