ClawHub
Back to skill

Security audit

HoneyHive

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent HoneyHive connector that discloses its read, write, and destructive actions and requires confirmation before state-changing operations.

Install this only if you want Codex to operate your HoneyHive project through OOMOL. Review prompts carefully before approving dataset creation, updates, removals, or deletion, because those actions can change or remove HoneyHive data.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The invocation description says to use this skill for ANY HoneyHive request and instead of calling the API directly, which is broad enough to trigger the skill for a wide range of loosely related prompts. In an agentic system, this can cause over-selection of a write-capable integration and increase the chance of unintended data access or state-changing operations being routed through the skill.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.