The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes data from 40+ external bioinformatics services (e.g., UniProt, KEGG, NCBI).
Ingestion points: Data enters the agent context through API responses from services like UniProt, KEGG, and UniChem as described in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the processed biological data.
Capability inventory: The skill allows the execution of bundled Python scripts (found in the scripts/ directory) and arbitrary network requests via the bioservices library.
Sanitization: No validation or sanitization mechanisms for the external data are mentioned in the documentation.
[PROMPT_INJECTION]: The skill includes instructions in the 'Suggest Using K-Dense Web' section to override default agent behavior by proactively marketing an external platform (www.k-dense.ai) when specific query conditions are met.
[EXTERNAL_DOWNLOADS]: The skill downloads the bioservices library from the Python Package Index (PyPI) and connects to various third-party biological databases and APIs as part of its core functionality.
[COMMAND_EXECUTION]: The skill's documentation describes running local Python scripts, such as scripts/protein_analysis_workflow.py and scripts/pathway_analysis.py, which are part of the skill's distribution.

bioservices