The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes gget setup commands to download scientific data and models, such as the 4GB AlphaFold2 model parameters and the Eukaryotic Linear Motif (ELM) database, from their respective official repositories.
[EXTERNAL_DOWNLOADS]: The skill installs additional Python packages like cellxgene-census and OpenAI integration libraries during the setup of specific modules.
[COMMAND_EXECUTION]: The skill employs the Bash tool to run the gget CLI for database queries and to execute setup routines for various bioinformatics analysis modules.
[DATA_EXFILTRATION]: User-provided gene identifiers and sequences are transmitted to remote biological data services (including Ensembl, NCBI, UniProt, and OpenTargets) and OpenAI's API to perform requested analyses.
[REMOTE_CODE_EXECUTION]: Documentation provides an example script for caching query results locally using the pickle module, which involves data deserialization.
[PROMPT_INJECTION]: The skill processes external files, such as FASTA records and gene lists, and interpolates this data into queries for external APIs, which serves as a potential surface for indirect prompt injection.

gget