cardiology-trial-editorial
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script, 'scripts/score_trial.py', to calculate clinical importance scores for cardiology trials based on metadata extracted by the agent.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from PubMed to generate cardiology editorials and HTML infographics. 1. Ingestion points: Data is retrieved via PubMed tools such as 'PubMed:search_articles' and 'PubMed:get_article_metadata'. 2. Boundary markers: No explicit instructions are provided to the agent to treat PubMed content as untrusted or to use delimiters. 3. Capability inventory: The skill generates HTML files and executes bundled Python scripts. 4. Sanitization: The instructions do not specify sanitization or escaping of the retrieved medical content before it is included in generated text or HTML documents.
- [DYNAMIC_EXECUTION]: The skill performs runtime execution of the 'scripts/score_trial.py' script as a standard part of its trial evaluation workflow.
Audit Metadata