Ae4
Medium
- Category
- analysis-evasion
- Confidence
- 80% confidence
- Finding
Suspicious Unicode normalization or mixed-script content
- Content
Security audit
Homemade Machine Learning Skill
Security checks for vulnerabilities and agentic risk
Overview
This skill is a coherent educational ML helper that prints explanations, code snippets, and notebook links without hidden persistence or sensitive access.
Install only if you are comfortable with an educational shell helper that prints static ML explanations and links to external notebook/source sites. Opening Binder or notebook links may involve third-party hosted content, but the packaged script itself does not fetch or execute remote code.
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
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- 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 (2)
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 88% confidence
- Finding
The skill description and command labels switch from English to Chinese for key usage instructions such as explaining algorithms, getting notebook links, and generating learning paths. This may violate language/locale policy because the skill imposes a specific language for operational guidance without user opt-in or justification.
- Content