skills/freedomintelligence/openclaw-medical-skills/tooluniverse-drug-drug-interaction/Gen Agent Trust Hub
tooluniverse-drug-drug-interaction
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill includes Python implementation scripts (
ddi_pipeline.py,ddi_working_example.py) designed to automate the data collection and report generation process using the ToolUniverse platform. These scripts perform standard API calls and file writing operations to generate markdown reports. - [DATA_EXPOSURE]: The skill creates and modifies local markdown files (e.g.,
DDI_risk_report_*.md) to store the results of the drug interaction analysis. This is a core part of its functionality and does not involve the exposure of sensitive system data or user credentials. - [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided drug names and incorporates them into report content and filenames.
- Ingestion points: User-supplied drug names in the prompt.
- Boundary markers: None explicitly implemented in the file-writing logic, though the context is narrow (clinical reporting).
- Capability inventory: File system write access (
open(),write()) and network access via authorized tools (RxNorm,DrugBank,DailyMed,PubMed,FAERS). - Sanitization: There is no explicit sanitization of drug names before they are used in filenames or report headers, which presents a theoretical low-risk surface for path manipulation if the environment is not properly sandboxed.
Audit Metadata