Src

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only setup guide for routing Codex through a local LiteLLM proxy to Vertex AI, with persistent configuration steps that are disclosed and purpose-aligned.

Install this only if you intend to route Codex through LiteLLM to Vertex AI. Use least-privileged Google Cloud credentials, apply trusted-workspace settings only to repositories you mean to trust, and consider setting the dummy OPENAI_API_KEY only for this workflow instead of globally in your shell profile.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Low
Confidence
93% confidence
Finding
The skill instructs users to add a credential-related environment variable to a persistent shell profile without warning that this change will affect future sessions. Even though the value shown is a placeholder for a local proxy, modifying shell startup files creates lasting side effects and can normalize unsafe handling of API key material in shared shells, logs, or inherited subprocess environments.

Session Persistence

Medium
Category
Rogue Agent
Content
## 1. LiteLLM Configuration

You need to create a `config.yaml` for LiteLLM that drops complex parameters and sets content to simple strings, then routes a `codex` model alias to your Gemini Vertex endpoint.

Create or update your `config.yaml` (e.g., `/app/config.yaml`):
Confidence
77% confidence
Finding
create a `config.yaml` for LiteLLM that drops complex parameters and sets content to simple strings, then routes a `codex` model alias to your Gemini Vertex endpoint. Create or update your `config.ya

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal