codex-computer-use

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These are direct installer scripts (install.sh/install.ps1) hosted on chatgpt.com — a non-official/typosquat-sounding domain — and the README’s curl | sh / iex pattern is a high-risk distribution method for malware.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow passes the caller-supplied --task free text directly into the Codex sub-agent prompt (buildPrompt() → createCodexArgs() → codex exec ... <prompt>), so if that task text contains outsider-authored content, it becomes LLM-readable context (indirect prompt injection risk).