Docs

MCP server

Connect Claude Code, Claude Desktop or Cursor in one step. The assistant sees what the person who connected it may see.

InsightClickIQ is an MCP server. Any client that speaks the Model Context Protocol can query a customer’s analytics through the same engine and tool registry the console’s agents use.

Endpoint and transport


URL	`POST https://clickiq.example.com/mcp`
Transport	Streamable HTTP, stateless, JSON responses: one JSON-RPC 2.0 request per POST, one JSON answer; no session ids, no SSE stream. `initialize` is optional
Protocol versions	2024-11-05, 2025-03-26, 2025-06-18, 2025-11-25 and 2026-07-28
Auth	`Authorization: Bearer <access token or API key>`. Create an API key under Users → API keys, or `POST /api/user/api-keys`
Scope	Every tool runs as the caller, pinned to their organisation by the same row-level security as the console; write tools need an editor

Claude Code

claude mcp add --transport http clickiq https://clickiq.example.com/mcp --header "Authorization: Bearer <token>"

Claude Desktop and Cursor

claude_desktop_config.json / .cursor/mcp.json

{ "mcpServers": { "clickiq": {
    "type": "http",
    "url": "https://clickiq.example.com/mcp",
    "headers": { "Authorization": "Bearer <token>" } } } }

curl

curl -s https://clickiq.example.com/mcp -H 'Authorization: Bearer <token>' \
  -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Tools

37 tools from the shared registry. Inputs are JSON schemas; results are compact JSON with table rows capped at 50 and every rate a fraction.

Tool	What it returns
`list_projects`, `get_catalog`	Projects; a project’s events (with 30-day counts), properties and samples, top pages, sources, countries
`run_query`	Any QuerySpec’s result; a validation error is the tool error
`compare_periods`, `drivers`	Now against the previous period; which source, referrer, device, country, browser or path carries a change
`web_overview`	The Overview’s KPIs, series, top pages and sources
`ask`	Ask’s answer with its queries, sessions and dashboards
`search_sessions`, `summarize_session`	Sessions by any filter (up to 25); a session’s intent, outcome, frustration and highlights
`list_signals`	Signals with impact and recommendation
`list_dashboards`, `get_dashboard`, `list_reports`, `render_report`	Dashboards with widget specs; reports rendered with their narrative
`journey_patterns`, `journey_compare`, `journey_dropoff`	Journeys
`heatmap_pages`, `heatmap_zones`, `explain_page`	Zones and the page explanation
`forms_list`, `form_detail`, `errors`, `error_detail`, `performance`	Forms and quality
`list_segments`, `segment_preview`, `list_annotations`	Audiences and markers
`list_flags`, `experiment_results`	Flags and experiments; an experiment’s variants, uplift, p, P(beat control) and readout
`list_surveys`, `survey_results`	Surveys with response rates and NPS; results and AI themes (counts and themes, never the raw open-text answers)
Editors only	`investigate`, `create_dashboard`, `save_segment`, `create_annotation`, `create_alert`