The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from the ClinPGx and PharmVar external databases. Because the agent processes these external responses to provide medical and genomic guidance, there is an inherent risk of indirect prompt injection if the external data source were to contain malicious instructions.
Ingestion points: Data is retrieved from the ClinPGx API (api.clinpgx.org) and PharmVar (pharmvar.org) through functions in scripts/query_clinpgx.py and the WebFetch tool.
Boundary markers: The skill documentation and helper scripts do not define explicit boundary markers or instructions to isolate the external data from the agent's core reasoning logic.
Capability inventory: The skill possesses the capability to execute Python scripts via Bash, read/write local files for data caching, and perform network operations.
Sanitization: The skill relies on standard JSON parsing (json.load and requests.json()) to process results, but does not implement specific filtering or sanitization to detect prompt-like content within the genomic annotations.

alterlab-clinpgx