polyhaven-texture-apply

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime path calls send('search_polyhaven_assets', ...) and send('download_polyhaven_asset', ...), which fetch PolyHaven asset data (outsider-authored public web content) and then passes the returned JSON fields (e.g., res.result.assets, res.result.maps, res.result.material) into the agent/LLM context via the tool/agent messaging layer.