The Agent Skills Directory

[SAFE]: The skill provides legitimate scientific data analysis functionality. All code snippets focus on numerical processing and visualization of simulation outputs.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool for standard simulation environment management via Conda and for extracting metadata from output files using grep. These operations are restricted to the intended analysis workflow.
[PROMPT_INJECTION]: While the skill ingests external simulation log files, the risk of indirect prompt injection is mitigated by the parsing logic, which primarily converts data to numerical types (integers and floats) before analysis, preventing the interpretation of embedded text as instructions.

data-analysis