math-pure-proofs
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: Analysis of the skill instructions and metadata reveals no malicious patterns, obfuscation, or unauthorized access attempts. The skill is entirely focused on logical reasoning and text generation within a sandbox environment.
- [INDIRECT_PROMPT_INJECTION]: The skill is subject to an indirect prompt injection surface as it processes untrusted mathematical statements. 1. Ingestion points: The skill accepts a mathematical statement as input (SKILL.md). 2. Boundary markers: None are specified to delimit the input from instructions. 3. Capability inventory: The skill is restricted to generating a Markdown artifact; it has no network, shell, or elevated file system access. 4. Sanitization: No sanitization of the input statement is described. Given the lack of dangerous tools, this category is assessed as safe.
Audit Metadata