bundle-social-uploads

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly ingests files from arbitrary public URLs via the POST /api/v1/upload/from-url endpoint (see references/05-upload.md and the SKILL.md "Import from web" pattern), so untrusted third‑party content can be fetched server‑side and then used by the agent (upload ID → used in subsequent post actions), enabling indirect injection to influence later behavior.