ClawHub
Back to skill

Security audit

Codex Review

Security checks across malware telemetry and agentic risk

Overview

This skill locally reviews Codex usage history and produces a local report, with sensitive access disclosed and no evidence of upload, deletion, or hidden behavior.

Install only if you are comfortable with a local skill reading recent Codex history and workspace artifact metadata to summarize your usage. Reports stay local according to the artifact, but they may still reveal project names, relative file paths, usage patterns, and derived prompt-quality observations to anyone who can open the generated files.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill enables implicit invocation without any trigger constraints, which can cause it to activate in broader contexts than the user explicitly intended. Because this skill scans local Codex sessions, project artifacts, skills, automation tasks, and token records, accidental invocation could expose sensitive local metadata or generate reports from private data without sufficiently clear user intent.

Missing User Warnings

Low
Confidence
91% confidence
Finding
This module aggregates sensitive local telemetry including session userMessages, titles, cwd paths, and artifact paths to infer project identity and produce reports. In the context of a local review skill, that collection may be expected, but there is no evidence in this file of minimization, redaction, consent gating, or output controls, so the code can expose private prompts, filesystem structure, project names, and work history if invoked on unintended data or if downstream reporting renders the collected fields.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
scripts/build_report.mjs:378