avoid-numeric-index
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [NO_CODE] (SAFE): The skill primarily consists of markdown documentation and TypeScript code examples for educational purposes. It does not implement any logic that executes system commands or handles external data.
- [PROMPT_INJECTION] (SAFE): Analysis of the markdown content found no attempts to override agent instructions, bypass safety filters, or extract system prompts. The 'Iron Rule' section is a stylistic guideline for code development, not an instruction for the AI agent to ignore its core safety parameters.
- [DATA_EXFILTRATION] (SAFE): No network requests, file system access, or credential markers were detected in the provided code snippets.
- [INDIRECT_PROMPT_INJECTION] (SAFE): This skill does not ingest untrusted external data and has no capability to perform write operations or external communication, presenting no surface for injection attacks.
Audit Metadata