The Agent Skills Directory

[SAFE]: The skill implements a comprehensive suite of simulation tools with a focus on reproducibility and safety. All bundled scripts include strict input validation logic located in scripts/_common.py that rejects network URLs, symlinks, and files exceeding size limits.
[SAFE]: Execution safety is enforced through a BoundedEnvironment which wraps the core SimPy engine to provide hard caps on the number of events, entities, and simulation duration, preventing non-terminating loops or resource exhaustion.
[SAFE]: The skill uses deterministic random stream separation and records all seeds in its output reports, ensuring that simulation results are verifiable and reproducible without introducing stochastic vulnerabilities.
[SAFE]: File system operations are handled securely; output files are written atomically using temporary files and restricted to owner-only permissions (0o600) to prevent unauthorized access or race conditions.
[SAFE]: The skill's documentation provides extensive guidance on simulation methodology and validation, referencing official sources from trusted organizations and well-known academic services such as PyPI, GitLab, and arXiv.

simpy