AI integration validation & protocol research

MNTN Public MCP Validation

Validation and integration research for customer-facing AI access to MNTN campaign and reporting capabilities.

View the workflow Explore the build

Business problem

The challenge

Exposing an advertising platform to AI assistants requires understandable tools and authorization that follows the caller.

Technical solution

Our approach

External AI client, public MCP server and authenticated PAPI requests.

  1. 01

    Validate the public interface

    Worked with and validated the customer-facing MCP for advertiser, campaign, flight, audience and reporting access.

  2. 02

    Understand action and access boundaries

    Researched how campaign lifecycle actions and caller authorization flow through the MCP server into PAPI.

  3. 03

    Evaluate future product integration

    Explored the relationship between external MCP tools and Nova assistant workflows, whose prototype currently uses direct PAPI tools.

Results & impact

Project outcomes

Validation and integration research for customer-facing AI access to MNTN campaign and reporting capabilities.

Validation

Public AI integration contribution

Contribution covers working with the MCP and understanding its platform integration.

Caller-scoped

Authorization model research

Requests retain the caller identity when forwarded to platform APIs.

Visual assets

How the workflow fits together

MNTN Public MCP Validation

flowchart TD
  A["External AI assistant"] --> B["Public MCP tools"]
  B --> C["Caller authorization"]
  C --> D["PAPI"]
  D --> E["Campaign and reporting data"]
01 External AI client, public MCP server and authenticated PAPI requests.

Tech stack

Technology and ownership.

Validation and integration research contribution. Authorship of the public MCP server is not part of this project claim.

Model Context ProtocolAI assistantsRESTPAPIAuthorization