ClawHub
Back to skill

Security audit

kr-patent-claims-review

Security checks for vulnerabilities and agentic risk

Overview

This skill is purpose-built for patent claim review, but it can automatically send sensitive patent claim details to PatSnap without a clear opt-in step.

Before installing, users should understand that patent claims or extracted technical features may be used in PatSnap MCP searches. Use this only with documents you are permitted to process through that service, and require explicit confirmation before any external search for confidential or unpublished inventions.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill instructs the agent to perform mandatory PatSnap searches based on claim content, but it does not clearly disclose before execution that user-uploaded patent material or extracted claim features may be transmitted to an external third-party service. In a patent-drafting context this is particularly sensitive, because unpublished claims can contain confidential invention details and undisclosed filing strategy, creating confidentiality, privacy, and commercial-risk exposure if sent externally without explicit informed consent.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger conditions are broad enough to match routine file-review requests such as 'help me check this document,' which can cause the skill to activate without the user clearly intending a Korea-patent-claims review workflow. Because the skill may then parse uploaded documents and proceed into legal/patent analysis steps, unintended invocation can expose sensitive content to unnecessary processing and trigger downstream tool use the user did not explicitly request.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.