Back to skill

Security audit

免费版

Security checks for vulnerabilities and agentic risk

Overview

The skill is a generic OpenRouter/SkillHub AI helper, but it asks for broad read, write, execute, API, and credential authority without clear limits or user controls.

Install only if you are comfortable giving this skill broad agent authority and potentially sending prompts or data to OpenRouter or downstream model providers. Avoid using it with secrets, regulated data, private files, or commands unless you add clear local controls for allowed files, commands, API endpoints, and explicit confirmation before external calls or writes.

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
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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation guidance is overly broad, encouraging use for generic AI model calls, conversations, orchestration, and workflow scenarios without narrow boundaries or decision criteria. In a skill with read/exec/write capabilities and external API usage, vague invocation conditions increase the chance the agent is triggered in inappropriate contexts and performs actions the user did not clearly intend.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The description indicates AI model/API usage but does not clearly warn users that their prompts or data may be transmitted to external services such as OpenRouter or provider-backed LLM endpoints. This is dangerous because users may unknowingly expose proprietary, personal, or regulated data to third parties under unclear handling and retention conditions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill advertises read, exec, and write capabilities without a clear warning that it can perform system-impacting actions. In an agent environment, this materially increases risk because users may invoke the skill expecting only model access while the skill also has the ability to modify files or execute commands.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The table at L045 says 'Free RideSkillHub管理' is not supported in the free version, but the manifest/description at L007-L017 says the skill 'manages free AI models from OpenRouter for SkillHub,' and the scenarios table at L055 also lists management operations as an output. These statements conflict about whether management capability exists in this skill.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill claims encrypted transmission and storage without any concrete implementation details, configuration, or verifiable controls in the file. This can create a false sense of safety, leading users to submit sensitive data under assumptions that may be untrue, especially because the skill appears to call external AI/API services.

Content

No source excerpt is available for this finding.