verify-quality-code-quality
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a defensive code review process designed to identify and block security vulnerabilities like XSS, SQL injection, and hardcoded credentials. It requires auditors to prioritize security and correctness findings over stylistic preferences.
- [PROMPT_INJECTION]: As a tool for auditing external code, the skill is exposed to potential indirect prompt injection through the data it processes.
- Ingestion points: User-provided code snippets or files referenced during the review process.
- Boundary markers: None explicitly defined in the skill instructions to delimit untrusted code from instructions.
- Capability inventory: The skill is limited to generating text-based review reports and does not have the capability to execute shell commands, make network requests, or modify system configurations.
- Sanitization: The skill relies on the underlying LLM's safety guardrails and does not specify custom sanitization for reviewed content.
Audit Metadata