Back to skill

Security audit

Hyperliquid CLI

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill benign with medium confidence: reported behavior appears limited, but the scanners flagged update and additional trading-related capabilities that are not clearly disclosed.

The verdict is benign, but confidence is medium and Skillspector's analysis is partial because a local path-like reference could not be resolved. Before installing, consider whether CLI updating and the additional dry-run operations fit your needs, and review any referenced material that is available.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The sole command requests a CLI update rather than any of the described trading functions. As the entire supplied implementation, this is a materially different primary operation, not merely a supporting detail within a demonstrated trading workflow.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
90% confidence
Finding

Vault deposit operations match the declared vault-management purpose. Staking delegation is a separate capability not disclosed in the description, and builder fee authorization is also not explicitly represented. Although the description mentions dry-run mutations, that describes an execution mode rather than clearly declaring these additional operation types. The mismatch is limited to capability scope; the chunk does not actually delegate funds, deposit assets, or grant an approval because all mutation commands use --dry-run.

Content

No source excerpt is available for this finding.