layer2

Warn

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [Unverifiable Dependencies & Remote Code Execution] (MEDIUM): The skill contains an explicit instruction for the agent to fetch content from an untrusted external URL (https://ethskills.com/l2s/SKILL.md). Since the target is another skill definition file, this facilitates the execution of unvetted remote logic.
  • [Dynamic Execution] (MEDIUM): By redirecting the agent to a remote resource for its core instructions, the skill allows for behavior that can be modified after static analysis, effectively bypassing local security controls.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 19, 2026, 07:10 PM