MCP prompts and JSON-RPC examples

Prompts#

• "Which pages on example.com got the most visitors from AI assistants last month, and how did that change from the month before?"
• "Compare signup conversion rate by channel for the last 90 days."
• "Create a goal for visits to /thank-you on example.com." (needs content.write)
• "Add a note on today's date: 'Launched new pricing page'."
• "Are any of my uptime checks down right now?"
• "How do I track outbound links with Privatus?" (uses docs_search)

The agent picks the tools, for example stats_breakdown with dimension=page, period=last_month, compare=previous_period and filters=[["channel","is","ai_assistants"]].

Raw JSON-RPC#

Useful for checking a token or building your own client.

List tools:

curl https://privatusanalytics.com/mcp \
  -H "Authorization: Bearer $PRIVATUS_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Call a tool:

curl https://privatusanalytics.com/mcp \
  -H "Authorization: Bearer $PRIVATUS_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{
    "jsonrpc": "2.0", "id": 2, "method": "tools/call",
    "params": {
      "name": "stats_aggregate",
      "arguments": { "site_id": "pa_7Q2K9XH3AB", "period": "7d", "compare": "previous_period" }
    }
  }'

The result's content[0].text is the same JSON the API returns ({"data": …}), also provided as structuredContent. Errors come back with isError: true and the API's error body.

Good practice#

• Use a dedicated token per agent, with the smallest set of permissions and sites, and an expiry date.
• Write tools (create, update, delete) show up only when the token has the permission. Keep agents on analytics.read unless they need more.
• Everything an agent changes is in the audit log under the token's member.