llm-eval-type-selector
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a standard attack surface for indirect prompt injection as it ingests and processes untrusted data.
- Ingestion points: The skill reads user-provided failure modes, quality dimensions, and potentially local project documentation such as
CLAUDE.mdorproduct-marketing-context.md(as directed in the "Before starting" section). - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the ingested data.
- Capability inventory: The skill contains no automated subprocess calls, file-write operations, or network tools, significantly limiting the impact of any potential injection.
- Sanitization: No specific filtering or sanitization is performed on the ingested text before it is processed by the agent.
- [SAFE]: The skill primarily provides natural language instructions and structural templates for the agent to assist a developer. No evidence of obfuscation, credential harvesting, or unauthorized remote code execution was found.
Audit Metadata