The Agent Skills Directory

[SAFE]: The skill follows a standard interactive workflow for game design. It uses a series of prompts and tool calls (AskUserQuestion, Write, Read, Task) to facilitate a collaborative brainstorming process between the AI and the user.
[DATA_EXPOSURE_AND_EXFILTRATION]: No sensitive file access or hardcoded credentials were found. The skill reads and writes within its expected project directories (design/gdd/, production/, and ../../docs/).
[REMOTE_CODE_EXECUTION]: There are no patterns of remote code execution or untrusted dependency installation. The skill references other agent tasks (creative-director, art-director, etc.) which are internal to the platform's orchestration system.
[PROMPT_INJECTION]: The instructions focus on creative facilitation and professional studio principles. There are no attempts to bypass safety filters or override system-level instructions.
[DYNAMIC_CONTEXT_INJECTION]: The skill does not use the !command syntax or any other form of silent shell execution at load time.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests existing design files and user input, these are part of the intended creative workflow. The skill uses standard interpolation for user answers without exposing high-privilege capabilities to untrusted data.

brainstorm