vercel-blob

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

• Third-party content exposure detected (high risk: 0.70). The skill's "List Files" and "Download File" examples (e.g., list({ prefix: 'uploads/' }) and fetching https://xyz.public.blob.vercel-storage.com/file.pdf) explicitly show the agent accessing public/user-uploaded blob URLs, which are untrusted third‑party user-generated content the agent may ingest.