naming-analyzer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes identifier names and code structures from user-provided files, which serves as an ingestion point for untrusted data.
- Ingestion points: The instructions specify analyzing variables, functions, classes, and other code identifiers provided by the user within the code context (referenced in SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the code content as strictly non-executable data or to ignore instructions potentially embedded in code comments or strings.
- Capability inventory: The skill identifies naming issues and explicitly proposes the generation of refactoring scripts to apply suggested changes across the codebase (referenced in SKILL.md and README.md).
- Sanitization: The instructions do not include any validation or sanitization requirements for the code content being processed.
Audit Metadata