The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill implements a workflow where the agent generates a JavaScript specification file (spec.js) and then executes it using a provided script (scripts/check.mjs). This script uses the Node.js vm module to run the simulation engine (runtime/des.js) against the generated spec. This is a core feature of the skill used to validate the model's logic and calculate sizing before building the final visualization. Because the execution is restricted to locally generated or skill-provided code, it represents a standard architectural pattern for this use case.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user descriptions (e.g., "a line that makes 2,000 motors") and interpolates this text into the spec.js file as a prompt field. This creates a surface for indirect prompt injection if a user attempts to include code-breaking characters or malicious instructions. However, the risk is mitigated by the agent's role in constructing the structured JavaScript file and the limited scope of the local simulation engine.

factory