ClawHub
Back to skill

Security audit

OceanToken Models

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed OceanToken model-pricing and delegation helper with appropriate warnings before sending user content to other models.

Install this only if you want OceanToken connected to your agent for model search, price estimates, and optional delegation. Before using delegation, confirm which model will receive your content, avoid secrets or private files, and watch cost estimates and account balance.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill description contains broad trigger phrases like 'which model to use', 'second opinion', and generic bulk text tasks such as summaries, translation, classification, and docs. In an agentic environment, these phrases can cause the skill to activate for many common user requests and route content to external OceanToken tools or third-party models when the request was not specifically about OceanToken, increasing the chance of unnecessary data sharing or unexpected tool use.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill is configured for implicit invocation and uses a very broad default prompt describing generic model selection, pricing, cost estimation, and delegation behaviors. This can cause the agent to trigger the skill in situations the user did not explicitly intend, potentially sending prompts or subtasks to external models and expanding data exposure or causing unintended spend.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.