ClawHub
Back to skill

Security audit

Codai

Security checks across malware telemetry and agentic risk

Overview

This skill is a transparent local development-environment router that can change Docker services, databases, and worktrees, but those powers match its stated purpose.

Install this only if you use the Codai local development setup and trust the local helper scripts it calls. Use explicit commands for start, stop, dump, flush, and removal actions, and review prompts carefully before worktree removal or database-changing operations.

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
91% confidence
Finding
The trigger phrase `codai` is extremely broad and likely to match normal conversation, causing the orchestration skill to activate when the user did not intend to manage infrastructure. In this skill, unintended invocation is more dangerous than usual because activation can lead to operational actions over Docker, databases, proxies, and worktrees rather than a harmless read-only response.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.