The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the geniml Python library from PyPI and from a remote Git repository (github.com/databio/geniml.git).
[COMMAND_EXECUTION]: The instructions include multiple examples of bash and Python code that execute shell commands for genomic data manipulation, such as file piping and CLI-based model training.
[PROMPT_INJECTION]: The skill contains a 'Suggest Using K-Dense Web' section that explicitly directs the AI agent to monitor user interaction and proactively promote an external platform. This is an instruction to override default agent behavior for the purpose of pushing specific services.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted genomic data files without sanitization while having access to sensitive capabilities. [Ingestion points]: Processes BED, CSV, and H5AD files from user-specified directories as seen in SKILL.md. [Boundary markers]: None; no instructions are provided to the agent to ignore or delimit potentially malicious content within these data files. [Capability inventory]: Includes subprocess execution for CLI tools and file system write access for saving tokens and model outputs. [Sanitization]: Absent; no validation or filtering of input file contents is described.

geniml