mathmodel-to-ei
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python initialization script (
scripts/init_manuscript.py) and standard LaTeX compilation tools (latexmk). The initialization script includes integrity checks using SHA-256 hashes to verify the bundled IEEE templates before copying them to the output directory. - [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process untrusted external data in the form of mathematical modeling papers (PDF, Word, or LaTeX formats). This presents a vulnerability surface for indirect prompt injection.
- Ingestion points: User-provided competition papers and project files (specified in
SKILL.md). - Boundary markers: The instructions do not define explicit boundary markers or delimiters to separate untrusted paper content from the agent's instructions.
- Capability inventory: The agent has capabilities for file system writes, execution of a local Python script, and invocation of LaTeX compilers.
- Sanitization: There is no mention of sanitization, filtering, or escaping of the content extracted from user papers before it is processed or used in project generation.
Audit Metadata