codex-review

Warn

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: MEDIUMPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to review source code, creating a surface for indirect prompt injection where malicious content in the code could influence agent behavior.
  • Ingestion points: Reads external source code and refactoring diffs for review (SKILL.md).
  • Boundary markers: No delimiters or instructions to ignore embedded instructions are specified for handling the untrusted input.
  • Capability inventory: The skill description identifies 'auto CHANGELOG generation' as a feature, which requires file system write access to project files.
  • Sanitization: No input validation, filtering, or escaping of external content is described.
  • [EXTERNAL_DOWNLOADS]: The installation instructions direct users to download the skill from an unverified GitHub repository ('BenedictKing/codex-review') that does not belong to a recognized trusted organization.
  • [COMMAND_EXECUTION]: The skill requires the 'Codex CLI' to be installed and executed on the host system, which involves running external software within the agent's execution environment.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 15, 2026, 06:10 PM
Security Audit — agent-trust-hub — codex-review