The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the user to execute scripts/generate_inputs.py to prepare computational workflows. Running this script executes local Python logic that triggers the creation of complex materials science data structures.
[REMOTE_CODE_EXECUTION]: The script scripts/generate_inputs.py imports jobflow_remote.submit_flow and contains code to transmit the generated computational DAG to a remote worker (perlmutter_worker) on an external HPC cluster. This behavior offloads local intent to remote infrastructure without explicit confirmation from the instruction set.
[REMOTE_CODE_EXECUTION]: The skill uses the jobflow-remote package, which is not listed in the environment constraints section of the skill documentation, representing an unverified dependency that facilitates remote interactions.
[COMMAND_EXECUTION]: Instructions in SKILL.md claim that the workflow generation script 'simply serializes the theoretical DAG to JSON'. However, the script code performs an active remote submission and fails to implement any file-writing logic for the --output parameter, misleading the user about the script's side effects and execution boundary.

mat-dft-electronic-transport