The Agent Skills Directory

[REMOTE_CODE_EXECUTION] (CRITICAL): The file references/api_reference.md contains a command to download and execute a shell script directly from a remote server.
Evidence: sh -c "$(curl -fsSL ftp://ftp.ncbi.nlm.nih.gov/entrez/entrezdirect/install-edirect.sh)".
Risk: This allows the remote server to execute arbitrary code on the user's system. The source ncbi.nlm.nih.gov is a legitimate institution but is not on the defined trusted source list for this analysis, and the execution pattern itself is inherently unsafe.
[EXTERNAL_DOWNLOADS] (LOW): The skill documentation provides numerous examples of interacting with external APIs via curl.
Evidence: Examples target https://eutils.ncbi.nlm.nih.gov/entrez/eutils/ for data retrieval.
Context: While expected for a ClinVar skill, these represent external network dependencies.
[NO_CODE] (INFO): The provided skill package is incomplete as it consists entirely of reference documentation.
Evidence: There is no SKILL.md file, no executable scripts, and no configuration files. It is a set of instructions/guides rather than a functional AI agent skill.

clinvar-database