Example use cases
- Audience and journey QA assistants that check configuration against rules.
- Implementation assistants that validate payloads against XDM schemas.
- Analytics copilots that answer questions from approved data views.
Governance principles
- Read-only access by default; write actions require approval.
- No secrets or personal data in prompts or logs.
- Every tool call logged and reviewable.
- Evaluation sets to measure accuracy before rollout.
Learn more
Read What is MCP?
Frequently asked questions
Is MCP an Adobe product?
No. MCP is an open protocol for connecting AI applications to tools and data. It can be used with enterprise systems where appropriate controls are in place.