The Agent Skills Directory

[SAFE]: A comprehensive security audit was performed across all 11 threat categories, including prompt injection, data exfiltration, and remote code execution. No malicious patterns or suspicious activities were identified. The skill's architectural patterns align with standard Godot development best practices.
[INDIRECT_PROMPT_INJECTION]: The skill uses Godot Resources to define card data. While these are ingestion points for external data, the implementation is safe as it does not involve interpolating this data into security-sensitive agent prompts or executing privileged commands based on the content.
[DYNAMIC_EXECUTION]: The architecture allows for dynamic instantiation of local Script resources to handle specific card logic. This is a standard and necessary pattern for data-driven game design and does not involve the execution of untrusted code from remote sources.

godot-genre-card-game