Skip to content

Quickstart

Connect quickly, then drop into deeper route detail only when you need it

The shortest path is to connect to the remote MCP, call one of the three flagship wrappers, and parse the returned decision-ready envelope.

Open capabilities View trust posture

Need full API route detail or response-contract guidance? Open the developer documentation hub.

Quickstart

Four steps to a first successful MCP call

  1. 1

    Discover the MCP

    Start from the registry listing or this overview page and confirm the server identity and transport.

  2. 2

    Connect to the remote server

    Use the live MCP endpoint at https://api.chinamarketing.ai/mcp over streamable-http.

  3. 3

    Make a paid request

    Use one flagship capability wrapper and let the x402 payment flow handle billing per request.

  4. 4

    Parse the response envelope

    Read the recommendation-style output and use the metadata for workflow routing, logging, or follow-up calls.

Server identity

ai.chinamarketing/intelligence

Remote endpoint

https://api.chinamarketing.ai/mcp

Transport

streamable-http

Capability examples

Short examples for the three flagship wrappers

This page stays intentionally short. The examples below are enough to show the shape clearly without turning `/mcp/docs` into a full route explorer.

Brand Visibility Snapshot

Diagnose where a brand is visible, weak, or missing across China-facing surfaces.

Returns summary, score, strengths, gaps, actions, and source cues.

Route details Agent framing

Example MCP call

json

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_brand_visibility_snapshot",
    "arguments": {
      "brand_name": "Chanel",
      "category": "makeup",
      "market_scope": "RED",
      "notes": "Prioritize discoverability and competitive momentum."
    }
  }
}

Example response

json

{
  "insight": "Chanel is showing rising China visibility momentum with strong RED discovery support.",
  "key_drivers": [
    "Trend momentum is improving for makeup-related discovery queries.",
    "Narrative tone remains positive across the latest sentiment pass."
  ],
  "recommendation": "Keep Chanel in the active monitoring set and compare against peer luxury beauty brands weekly.",
  "confidence": "medium",
  "source_summary": {
    "trend_lane": "qwen",
    "sentiment_lane": "qwen",
    "snapshot_support": "category-aligned"
  },
  "cost": {
    "amount_usd": "0.63",
    "billing_basis": "per_request"
  }
}

Destination Demand Snapshot

Read outbound China demand direction for one destination in one response.

Returns direction, notable signals, risks, opportunities, and actions.

Route details Agent framing

Example MCP call

json

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_destination_demand_snapshot",
    "arguments": {
      "destination": "Tokyo",
      "origin_market": "China",
      "travel_window": "2026-02",
      "audience_segment": "luxury travelers"
    }
  }
}

Example response

json

{
  "insight": "Tokyo is showing rising China outbound movement with supportive destination-spend context.",
  "key_drivers": [
    "Estimated passengers are up versus the prior month.",
    "Shanghai to Tokyo remains the top feeder route.",
    "Destination spend context remains supportive."
  ],
  "recommendation": "Keep Tokyo in the active destination monitoring set and compare it against peer short-haul markets.",
  "confidence": "medium",
  "source_summary": {
    "flight_month": "2026-02",
    "spend_source_month": "2026-02",
    "route_concentration": 0.34
  },
  "cost": {
    "amount_usd": "0.20",
    "billing_basis": "per_request"
  }
}

KOL Shortlist

Generate a creator shortlist with rationale, risk flags, and next steps.

Returns candidate creators, rationale, risks, and next actions.

Route details Agent framing

Example MCP call

json

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_kol_shortlist",
    "arguments": {
      "brand_name": "Chanel",
      "category": "beauty",
      "campaign_goal": "launch visibility",
      "platform_focus": ["red"]
    }
  }
}

Example response

json

{
  "insight": "Beauty creator discovery is strongest on RED with a compact high-fit shortlist.",
  "key_drivers": [
    "Current evidence favors mid-tier beauty creators with strong category fit.",
    "RED remains the clearest platform lane for this brief."
  ],
  "recommendation": "Use the shortlist as the starting set for campaign planning, then narrow further by tier or geography.",
  "confidence": "high",
  "source_summary": {
    "provider_lane": "proxy",
    "shortlist_count": 12
  },
  "cost": {
    "amount_usd": "0.45",
    "billing_basis": "per_request"
  }
}

Response model

Designed for decision-ready envelopes, not raw data dumps

Decision-ready outputs lead with `insight`, `key_drivers`, and `recommendation` so an agent can act without re-summarising the response first.

Responses expose `confidence`, `source_summary`, and `cost` so automation layers can reason about evidence strength and budget together.

Metadata is shaped for repeated workflow use instead of one-off reading.