codex-review

Warn

Audited by Gen Agent Trust Hub on Sep 30, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct the user to install a package from a third-party GitHub repository (BenedictKing/codex-review). This repository is not managed by the skill author (davila7) and does not belong to a recognized trusted organization.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to perform code reviews and generate changelogs, which involves ingesting untrusted source code from project files.
  • Ingestion points: The skill reads local project files for review (referenced in SKILL.md).
  • Boundary markers: There are no explicit instructions or delimiters defined to separate the code content from the agent's primary instructions.
  • Capability inventory: The skill performs code analysis and file modifications (changelog generation), implying read and write capabilities.
  • Sanitization: No evidence of sanitization or safety-filtering for the ingested code content is provided in the skill documentation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 30, 2026, 05:47 PM