taste-skill

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

• Third-party content exposure detected (high risk: 0.75). The required workflow is installing/invoking the upstream taste-skill from a public GitHub repo via npx skills add Leonxlnx/taste-skill ..., which at runtime can fetch outsider-authored free text (the repo’s README/SKILL content and any embedded instructions) into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's install step ("npx skills add Leonxlnx/taste-skill") fetches and executes code from the upstream repository https://github.com/Leonxlnx/taste-skill during installation, so remote content is required and can directly control the skill's behavior.