The Agent Skills Directory

[SAFE]: The skill instructions are focused on legitimate engineering practices for simulation software, including physics modeling, sensor simulation, and validation workflows. The guidance includes explicit boundaries between functional engineering and security testing, directing users to specialized security skills for hardware-in-the-loop exploitation or formal verification.
[INDIRECT_PROMPT_INJECTION]: The skill defines a scenario runner architecture that ingests external parameter sets and world configurations as described in references/sensors_environment_and_scenarios.md. While this represents a data ingestion surface, it is fundamental to the simulation engineering role and is mitigated by the skill's emphasis on deterministic execution and validation metrics.
Ingestion points: Scenario loading via run(scenario_id, params, seed) and Scenario DSL defined in references/sensors_environment_and_scenarios.md.
Boundary markers: The skill does not explicitly define delimiters for scenario files but emphasizes strict versioning and schema validation for parameter packs.
Capability inventory: The skill assumes standard capabilities for mathematical modeling, code generation, and system integration without requesting additional platform privileges.
Sanitization: The instructions include logic for NaN guards and safe defaults in references/integration_sil_hil_and_twins.md to handle malformed or erroneous data.

simulation-software-engineer