The Agent Skills Directory

[Remote Code Execution] (HIGH): The file references/api_reference.md contains a command to download and execute a shell script from a remote FTP server using sh -c "$(curl -fsSL ...)". This pattern is a major security risk as it can be used to execute arbitrary malicious code on the host system. While the source is a government institution (NCBI), it is not within the explicitly defined trusted organizations and the execution pattern itself is unsafe.
[Indirect Prompt Injection] (LOW): The skill facilitates the ingestion of external clinical variant data from the NCBI API.
Ingestion points: REST API responses from eutils.ncbi.nlm.nih.gov.
Boundary markers: Absent; the documentation does not suggest using delimiters to separate data from instructions.
Capability inventory: The skill uses curl and Python's Biopython library to fetch and process data, and it includes instructions for shell-based tool installation.
Sanitization: No mention of input validation or sanitization for the retrieved API data, which could contain malicious content targeting the downstream agent.

clinvar-database