The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a recipe for data persistence using the pickle module (pickle.load). Deserializing data from untrusted pickle files is a known security risk that can lead to arbitrary code execution.\n- [PROMPT_INJECTION]: The skill ingests external data from CSV files (counts.csv and metadata.csv). While this constitutes a data ingestion surface for indirect prompt injection, the risk is limited as the data is used for statistical modeling and visualization rather than being interpreted as instructions by the agent.\n- [EXTERNAL_DOWNLOADS]: The skill utilizes several standard Python packages including pydeseq2, matplotlib, and seaborn which are instructed to be installed via pip. These are well-known and trusted libraries within the bioinformatics and data science communities.\n- [SAFE]: No evidence of hardcoded credentials, malicious network requests, obfuscation, or persistence mechanisms was found in the analyzed skill file.

pydeseq2-differential-expression