The Agent Skills Directory

[SAFE]: The skill is designed for project management and documentation within a game development context. It uses standard file operations on project-specific paths like design/gdd/.
[PROMPT_INJECTION]: No evidence of instructions intended to bypass AI safety guardrails, override system prompts, or extract internal configurations was found.
[DATA_EXFILTRATION]: The skill does not perform network requests or access sensitive user credentials. All operations are confined to the local project workspace.
[REMOTE_CODE_EXECUTION]: There are no patterns involving the download of external scripts or the execution of untrusted code. It does not utilize package managers like npm or pip.
[COMMAND_EXECUTION]: The skill does not use shell execution tools or attempt to gain elevated privileges (e.g., sudo).
[INDIRECT_PROMPT_INJECTION]: The skill reads user-provided game concept files and passes content to other review agents. Although this represents an ingestion surface for untrusted data, the skill includes human-in-the-loop checkpoints and the output is limited to Markdown documentation, significantly minimizing the risk.

map-systems