ClawHub
Back to skill

Security audit

Near Best Practices

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local NEAR best-practices lookup guide with an overstated term count but no hidden access, persistence, or unsafe behavior found.

Users can treat this as a simple local reference tool, but should not rely on it as a comprehensive NEAR security guide because its advertised coverage is overstated and the guidance is limited.

A.I.G

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

SkillSpector

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 (1)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The code does align in broad theme with the declared purpose: it is a NEAR best-practices reference covering wallet security, smart contracts, DeFi, and related topics. It does not perform hidden actions, access external resources, or use undeclared permissions. However, the description materially overstates the scope: the script contains only roughly 21 entries, not 100+ terms, and functions mainly as a simple CLI lookup/browsing tool rather than a comprehensive guide. This is therefore a description-to-behavior mismatch based on overstated coverage and scope.

Static analysis

No suspicious patterns detected.