The Agent Skills Directory

Command Execution: The skill utilizes shell commands and Python execution during its internal verification phase. Specifically, it employs python3 -c to programmatically validate the mathematical properties of generated items. This ensures that the assessment's answer keys and distractors are mathematically consistent and accurate before they are presented to the user.
Indirect Prompt Injection Surface: The workflow involves ingesting data from external Knowledge Graph tools to retrieve standards, learning components, and misconceptions. While this ingestion process represents a potential surface for indirect prompt injection, the skill mitigates this risk through explicit instructions to synthesize original content and not reproduce retrieved text verbatim. These guardrails ensure that external data is used for grounding rather than direct output.

k12-check-for-understanding