ClawHub
Back to skill

Security audit

OpenAI Codex Multi OAuth

Security checks across malware telemetry and agentic risk

Overview

This is a coherent Codex profile debugging skill, but it exposes local account/session details and can use stored OAuth tokens to query usage data, so users should review it carefully before installing.

Install only if you intentionally need local OpenClaw Codex OAuth/profile diagnostics. Run scripts manually, prefer --profile or a test --state-dir to limit scope, avoid --raw unless necessary, and redact emails, account IDs, workspace IDs, session targets, and usage output before sharing logs or screenshots.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The guidance explicitly recommends displaying `user_id`, `account_id`, and `email` during debugging, which are sensitive identifiers that can expose personal or tenant information in chat logs, screenshots, bot outputs, or shared debug artifacts. In this skill’s context—multi-profile OAuth debugging across shared workspaces and routers—the chance of cross-account disclosure is higher, making routine exposure of these fields unnecessary and risky unless tightly limited.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The script prints sensitive local state including profile emails, session targets, and auth profile overrides directly to stdout with no masking, consent prompt, or warning. In debugging workflows this output can easily be copied into logs, terminal history, screenshots, or shared issue reports, creating unintended disclosure of personal and operational metadata.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.