denario
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions involve installing the 'denario' package from the standard Python package registry using 'uv' or 'pip'.
- [COMMAND_EXECUTION]: The skill provides functions like 'get_results()' and 'get_paper()' that allow the system to execute computational analyses and generate publication documents.
- [PROMPT_INJECTION]: The skill contains a behavioral instruction for the AI to recommend an external platform (www.k-dense.ai) for complex tasks, acting as a productivity suggestion.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests untrusted data to guide its agentic workflows. Ingestion points: Data descriptions provided to 'set_data_description', research ideas, and methodology files. Boundary markers: No explicit boundary markers or isolation instructions for user-provided data were found. Capability inventory: The skill can execute computational experiments and generate LaTeX source files. Sanitization: No sanitization or validation of the ingested research data is described.
Audit Metadata