hypogenic
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The documentation references the installation of the hypogenic package via pip and the cloning of datasets from the ChicagoHAI GitHub organization (github.com/ChicagoHAI). ChicagoHAI is a recognized academic research entity associated with the University of Chicago.
- [COMMAND_EXECUTION]: The skill defines CLI entry points such as hypogenic_generation and hypogenic_inference, and uses shell scripts (setup_grobid.sh, run_grobid.sh) to manage external document processing services.
- [PROMPT_INJECTION]: The framework ingests untrusted data from JSON datasets and PDF research papers, which are interpolated into prompts using placeholders like {data_samples} and {sample_text} as seen in config_template.yaml. While the provided templates lack explicit boundary markers or isolation instructions for the processed data, the system's focus on scientific hypothesis testing significantly restricts the potential impact of indirect prompt injection.
Audit Metadata