codex-review
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installation command
npx skills add -g BenedictKing/codex-reviewfetches content from a GitHub repository belonging to a third-party user not associated with the provided author context.- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves reading project source code to generate reviews and changelogs, creating a vector for instructions embedded in the data. - Ingestion points: Local source code files provided for review (referenced in
SKILL.md). - Boundary markers: No explicit delimiters or 'ignore' instructions are documented to protect the agent from embedded malicious prompts in code.
- Capability inventory: The skill reads local project files and has the capability to write to a
CHANGELOG.mdfile in the project root. - Sanitization: There is no mention of sanitizing or filtering the source code content before it is processed by the AI model.
Audit Metadata