Back to skill

Security audit

Smart Model Switching

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only routing guide for choosing Claude model tiers, with no code, install hooks, persistence, credential use, or hidden data handling.

Install this if you want model-selection guidance for Claude tier routing. Be aware it is Claude-specific and may bias agents toward cheaper models for simple work, so review important or high-stakes tasks if model quality matters.

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
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The line states the skill is 'Built for Claude-only setups,' which imposes a provider-specific constraint in natural language. The policy allows constraints when they are explicitly justified or optional, but this wording presents the restriction as mandatory without user opt-in or explanation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.