The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes Python modeling code that utilizes standard numerical and scientific libraries such as NumPy, Pandas, and SciPy. The code is strictly mathematical and does not involve arbitrary command execution, system-level calls, or unsafe subprocess operations.
[EXTERNAL_DOWNLOADS]: The references mention standard industry libraries like windpowerlib and scientific packages. There are no patterns suggesting the download of untrusted scripts or the execution of remote payloads.
[DATA_EXFILTRATION]: No network-enabled operations (e.g., curl, requests) or access to sensitive system paths (e.g., .ssh, .aws) were found. The skill operates entirely on provided input data for engineering calculations.
[PROMPT_INJECTION]: The instructions in SKILL.md guide the agent to prioritize the provided technical references for domain accuracy. There are no attempts to override safety filters, extract system prompts, or induce unrestricted behavior.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process weather datasets and load profiles. Although it handles external data, the processing is limited to mathematical simulations and data frame manipulation, lacking the capability tier required for successful indirect prompt injection attacks.

renewable-energy