tooluniverse-spatial-omics-analysis
Warn
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill explicitly instructs the agent to 'write and run Python code via Bash' when computation, statistics, or data processing is required. This practice involves generating executable content at runtime based on the analysis context.
- [COMMAND_EXECUTION]: The instructions mandate the use of a shell environment ('via Bash') to execute the dynamically generated Python scripts, which is a standard pattern for command execution in agentic workflows.
- [INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect injection as it ingests untrusted user input such as gene lists ('svgs'), tissue types, and disease context. These inputs are used to populate database queries (OpenTargets, HPA, STRING) and may be interpolated into the dynamically generated Python code.
- Ingestion points: User-provided parameters
svgs,disease_context,spatial_domains, andmetabolites(SKILL.md). - Boundary markers: The skill uses a structured report-first approach but does not explicitly instruct the agent to sanitize or delimit the external biological data before processing.
- Capability inventory: The skill has access to shell execution ('via Bash'), file writing (report generation), and a wide array of network-backed bioinformatics tools.
- Sanitization: There is no evidence of specific input validation or sanitization for the gene symbols or disease names provided by the user.
Audit Metadata