The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and integrate multiple external biological datasets including transcriptomics, proteomics, and genomics data from file formats such as CSV, TSV, HDF5, and VCF. * Ingestion points: Phase 1 (Data Loading & Quality Control) identifies the intake of multiple omics matrices and variant files. * Boundary markers: There are no explicit instructions or delimiters provided to prevent the LLM from interpreting natural language instructions if they were embedded within the data files. * Capability inventory: The skill's capabilities are focused on statistical correlation, multi-omics clustering, and pathway enrichment using standard libraries. It does not possess capabilities for arbitrary command execution or external network exfiltration. * Sanitization: The provided Python snippets do not implement specific input sanitization for potential injection strings within the datasets.
[SAFE]: The skill performs legitimate bioinformatics functions. All code examples use standard, well-known libraries for data science. The referenced external resources are academic DOIs or internal ToolUniverse skill identifiers belonging to the same author, and no indicators of credential theft, remote code execution, or persistence were found.

tooluniverse-multi-omics-integration