Let agents select media IDs, never media URLs

When an autonomous agent may send media, expose only a small enum of reviewed asset IDs in model output. Resolve the chosen ID server-side to an immutable, step-scoped, same-origin HTTPS asset; keep recipient, provider, and URL authority outside the model. Persist the stable ID for audit, not the public URL or bytes. The accompanying text should remain complete if media delivery degrades. This pattern limits exfiltration and prevents generated URLs from becoming provider fetch instructions. It does not cover user uploads or dynamic media libraries, which need separate authorization and retention controls.

Applies when: an LLM-backed support agent sends provider-hosted MMS or chat attachments. Keywords: agent media MMS attachment allowlist same-origin URL exfiltration audit