The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the bioservices package from the Python Package Index (PyPI) to function.
[COMMAND_EXECUTION]: The skill includes several Python scripts (batch_id_converter.py, compound_cross_reference.py, pathway_analysis.py, protein_analysis_workflow.py) designed to be executed via the command line to perform data retrieval, processing, and file writing operations.
[INDIRECT_PROMPT_INJECTION]: The skill processes biological data retrieved from external bioinformatics APIs, which could theoretically contain malicious instructions.
Ingestion points: Data enters the agent context through API responses from UniProt, KEGG, NCBI, PSICQUIC, and QuickGO, as seen in the retrieval methods across all scripts in the scripts/ directory.
Boundary markers: There are no explicit delimiters or specific instructions provided to the agent to disregard potential command-like patterns within the scientific data retrieved.
Capability inventory: The skill possesses network communication capabilities and the ability to write result files to the local file system (CSV, SIF, and TXT formats).
Sanitization: While the scripts perform structural parsing of the retrieved data (e.g., extracting specific fields from XML or TSV), there is no specific sanitization designed to filter out natural language instructions that might influence the agent's behavior.

bioservices