modeling-paper-rubric-and-model-selector
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed for academic paper planning and operates exclusively on local project files. Analysis of the scripts and instructions reveals no malicious behavior.
- [COMMAND_EXECUTION]: The skill utilizes local shell commands to invoke Python scripts for internal workflow coordination, such as
workflow_guard.pyandupdate_workflow_memory.py. These operations are limited to project-specific logic and do not involve external command injection or network interactions. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from a local
problem_analysis.jsonfile to generate paper structures. This represents standard data ingestion for its intended purpose and does not exhibit patterns associated with prompt injection attacks. - [DYNAMIC_EXECUTION]: The script
build_model_route.pyuses static mapping logic and templates to organize its output, with no evidence of dynamic code generation or runtime compilation.
Audit Metadata