game-assets

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). Runtime path skill-doc → _fetch_remote_skill_doc() fetches remote Markdown from .../api/agent-skills/game-assets/doc (outsider-authored guide content) and returns it as payload["content"], which the agent then uses in its LLM context for command selection.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The CLI bootstrap checks the remote manifest at https://raw.githubusercontent.com/Meowa-AI/meowa-skills/main/skills/game-assets/meowart_api.bootstrap.json and, if newer, downloads and executes the runner from https://raw.githubusercontent.com/Meowa-AI/meowa-skills/main/skills/game-assets/meowart_api.py at runtime (os.execve), enabling remote Python code to control execution.