think-problem-restatement
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to generate restatements, which is a surface for indirect prompt injection. Ingestion points: User-provided problem statements in SKILL.md. Boundary markers: None defined. Capability inventory: No tools, code execution, or network calls. Sanitization: None.
- [NO_CODE]: The skill is composed of markdown and metadata without any executable scripts.
- [SAFE]: The repository link for the author's collection is included as a reference.
Audit Metadata