code-quality-review
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data in the form of code diffs and repository context. This creates a vulnerability surface where adversarial instructions could be placed within code comments or strings to manipulate the agent's review outcome.
- Ingestion points: The agent reads current branch diffs and surrounding repository code (SKILL.md).
- Boundary markers: The instructions provide logical boundaries for the review but do not include technical delimiters or specific directives to ignore instructions embedded within the analyzed code.
- Capability inventory: The potential impact is minimized by the skill's configuration, which includes
disable-model-invocation: true, preventing the agent from calling external tools or performing actions beyond generating text analysis. - Sanitization: There is no evidence of sanitization, filtering, or escaping of the content found within the code diffs.
Audit Metadata