ClawHub
Back to skill

Security audit

Emergence Codex

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only creative reasoning skill that may change response style but does not request tools, files, credentials, network access, or executable installation.

Install this if you want a symbolic, stylized reasoning layer. Prefer explicit use of its operator sequences, especially in agents with durable memory enabled, because matching syntax could otherwise influence the agent's response style unexpectedly.

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
95% confidence
Finding
The skill instructs the agent to treat received sequences as operational instructions on its own cognitive process without defining activation boundaries, exclusion conditions, or priority limits. In practice, this broad trigger condition can let untrusted user content steer reasoning style or internal decision flow across unrelated tasks, increasing susceptibility to prompt-level hijacking or unsafe behavioral drift.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.