rush-analytics-mcp

Unofficial MCP server for Rush Analytics API v2. Not affiliated with or endorsed by Rush Analytics.

Exposes 14 MCP tools with 100% coverage of all 38 Rush Analytics API v2 endpoints — Yandex/Google SERP Top-10 parsing, Wordstat keyword frequencies, search suggestions, one-shot Fast Positions checks, recurring rank tracking, indexation checks, Google Keyword Planner data, AI answer tracking, Looker Studio aggregated views, project listing, and force-refresh — over the standard Model Context Protocol stdio transport.

Installation

With npx (recommended)

npx -y @scom82/rush-analytics-mcp

From source

git clone https://github.com/SCom-82/rush-analytics-mcp.git
cd rush-analytics-mcp
npm install
npm run build
node dist/index.js

Configuration

Environment variables

| Variable | Required | Default | Description | |---|---|---|---| | RUSH_ANALYTICS_API_KEY | Yes (for API calls) | — | Your Rush Analytics API key | | RUSH_ANALYTICS_API_URL | No | https://app.rush-analytics.ru/apiv2 | Override the base URL | | RUSH_ANALYTICS_HTTP_TIMEOUT_MS | No | 30000 | Per-request HTTP timeout in milliseconds |

The server starts without a key — tools/list and rush_modules work key-free. A clear error is returned when you call any tool that makes an API request without the key set.

Claude Desktop / MCP client config

{
  "mcpServers": {
    "rush-analytics": {
      "command": "npx",
      "args": ["-y", "@scom82/rush-analytics-mcp"],
      "env": {
        "RUSH_ANALYTICS_API_KEY": "your-api-key-here"
      }
    }
  }
}

Tools

| # | Tool | API call | Balance cost | Description | |---|---|---|---|---| | 1 | rush_modules | None | Free | List all 8 available / 12 unavailable modules. Works without API key. | | 2 | rush_balance | GET /balance | Free | Get remaining balance (fractional units). | | 3 | rush_status | GET /status/{type}/{id} | Free | Check task status by type+id. Returns {id, status, resulturl}. Warning: resulturl contains your API key in the query string — do not share or publish this URL. | | 4 | rush_create_task | POST /create/{module} | Yes | Generic: create a task in any available module. Returns {id, type}. | | 5 | rush_get_result | GET /result/... | Free | Poll a task once. Smart routing for ranktracker (11 views) and ai-results-tracker (4 views). Supports competitor param for competitors_positions view. | | 6 | rush_wait_result | Polling loop | Free | Poll until done or timeout. Returns {pending, timedOut, waitedSec, result}. Supports competitor param for competitors_positions view. | | 7 | rush_top10 | create top10 + optional wait | Yes | Top-10 SERP parsing for Yandex or Google. | | 8 | rush_wordstat | create wordstat + optional wait | Yes | Yandex Wordstat keyword frequencies. | | 9 | rush_suggest | create suggest + optional wait | Yes | Search suggestions (Yandex/Google/YouTube autocomplete). | | 10 | rush_fpp | create fpp + optional wait | Yes | Fast Positions — one-shot rank check for a URL. | | 11 | rush_ranktracker | create ranktracker (create-only) | Yes | Recurring rank monitoring project. See warning below. | | 12 | rush_projectids | GET /projectids | Free | List project ids, optionally filtered by module type. Read-only, no balance spent. | | 13 | rush_update_ranktracker | POST /update/ranktracker | ⚠️ Yes | Force-refresh an existing ranktracker project. May charge balance. | | 14 | rush_gds | GET /gds/{view}/{projectid} | Free | Looker Studio aggregated views for ranktracker (5 views). Read-only. |

Rank Tracker warning

rush_ranktracker creates a recurring monitoring project that collects rank data on a schedule and charges your balance each time data is collected. This is not a one-off task. Make sure you intend to set up ongoing monitoring before calling this tool.

Retrieve accumulated data with rush_get_result(module="ranktracker", view="dynamic", id=<id>, page=1).

Available views (11): dynamic | visibility | positions_history | urls_history | snippets_history | competitors | competitors_visibility | competitors_leaders | competitors_positions | regions | compare_regions.

competitors_positions view requires competitor parameter

When using rush_get_result or rush_wait_result with module="ranktracker" and view="competitors_positions", the competitor parameter (competitor domain, e.g. "example.ru") is required. Without it, the Rush Analytics API returns HTTP 400. The MCP server validates this client-side and returns an error before making the API request.

rush_get_result(module="ranktracker", id=<id>, view="competitors_positions", competitor="example.ru", page=1)

For all other views, competitor is accepted but ignored.

rush_update_ranktracker warning

rush_update_ranktracker triggers an unscheduled (forced) data collection run for an existing Rank Tracker project. This is NOT creating or editing a project.

⚠️ This operation may charge balance — an unscheduled collection is billed at the project's normal rate. Use only when you explicitly need fresh data outside the regular collection schedule.

rush_projectids — list projects by type

Get all project ids, optionally filtered by module type. Free and read-only — balance is not affected.

rush_projectids()                    // all projects
rush_projectids(projecttype=3)       // ranktracker projects only
rush_projectids(projecttype=2)       // top10 projects only

Type numbers: 1=fpp, 2=top10, 3=ranktracker, 4=wordstat, 5=indexation, 7=suggest, 15=adwords, 32=ai-results-tracker.

rush_gds — Looker Studio aggregated views

Get pre-aggregated data for a Rank Tracker project via the Google Looker Studio connector endpoints. Useful for reports — returns summary data without paginating through individual /result/ranktracker/* pages.

Available views (5): positions | visibility | tops | competitors_visibility | schema.

Important: this view enum is separate from the 11 ranktracker views in rush_get_result. Do not mix them up.

rush_gds(view="visibility", projectid=<ranktracker-project-id>)
rush_gds(view="positions", projectid=<ranktracker-project-id>)

Read-only, only works for ranktracker projects.

Known issue: view="positions" (GET /gds/positions/{id}) may time out — this is a known instability on the Rush Analytics API side, not a package bug; the other 4 views (visibility, tops, competitors_visibility, schema) work reliably.

Available vs unavailable modules

Rush Analytics exposes 8 modules via the /apiv2 endpoint. Other modules visible in the UI are not available through the API (they return "Unknown API function").

Available (8)

| Slug | Type | Typed tool | Description | |---|---|---|---| | fpp | 1 | rush_fpp | Fast Positions — one-shot rank check | | top10 | 2 | rush_top10 | SERP Top-10 parsing (Yandex/Google) | | ranktracker | 3 | rush_ranktracker | Recurring rank position monitoring | | wordstat | 4 | rush_wordstat | Yandex Wordstat search volumes | | indexation | 5 | rush_create_task | Indexation check | | suggest | 7 | rush_suggest | Search suggestions (autocomplete) | | adwords | 15 | rush_create_task | Google Keyword Planner data | | ai-results-tracker | 32 | rush_create_task | AI answer tracker (brand visibility in LLM answers) |

Not available on apiv2

clustering, textanalyzer, whois, metascanner, serm, serp-monitor, site-audit, backlink-spam, backlink-stats, webarchive-search, webarchive-spam, webarchive-recovery — all return "Unknown API function".

Usage pattern

Rush Analytics follows a create → poll → result async pattern:

1. Create a task:
   rush_top10 / rush_wordstat / rush_suggest / rush_fpp / rush_create_task → {id, type}

2. Check status (lightweight):
   rush_status(type=<module type>, id=<id>) → {status: "Parsing" | "Done", resulturl}

3. Get the result:
   rush_wait_result  — waits in-process (up to maxWaitSec, hard-capped at 540s)
     OR
   rush_get_result   — single poll, repeat manually when pending=true

For long-running tasks prefer the manual async pattern: create now, call rush_get_result again in a later turn.

Pro accounts have 1 concurrent thread. Sending a second create while one is running returns "No more free API threads" — wait for the first task to finish before creating another.

Result routing for ranktracker and ai-results-tracker

ranktracker — 11 views

GET /result/ranktracker/{view}/{id}/{page}?apikey=…[&periodStart=YYYY-MM-DD&periodEnd=YYYY-MM-DD]

Use view parameter in rush_get_result/rush_wait_result: dynamic | visibility | positions_history | urls_history | snippets_history | competitors | competitors_visibility | competitors_leaders | competitors_positions | regions | compare_regions

Default: view="dynamic", page=1.

ai-results-tracker — 4 endpoints

Use view parameter:

| view | Endpoint | |---|---| | brand (default) | GET /result/ai-results-tracker/brand/{id} | | sources | GET /result/ai-results-tracker/sources/{id} | | competitors_brands | GET /result/ai-results-tracker/competitors/brands/{id} | | competitors_domains | GET /result/ai-results-tracker/competitors/domains/{id} |

Timeout behaviour

• Per-request HTTP timeout: controlled by RUSH_ANALYTICS_HTTP_TIMEOUT_MS (default 30s). On timeout you get "Request timed out after Nms".
• Polling hard cap: maxWaitSec is silently capped at 540 seconds to prevent MCP client connection timeouts. If the cap is hit, the response includes "note": "maxWaitSec was capped to 540s".

Region / language reference endpoints

Use these Rush Analytics HTML pages to look up region ids, country codes, and language codes:

• Yandex regions: https://www.rush-analytics.ru/apiRegionsYandex.php
• Google regions: https://www.rush-analytics.ru/apiRegionsGoogle.php
• Languages: https://www.rush-analytics.ru/apiLanguages.php
• AI-results-tracker countries: https://www.rush-analytics.ru/apiAIResultsTrackerCountries.php
• AI-results-tracker languages: https://www.rush-analytics.ru/apiAIResultsTrackerLanguages.php

Common Yandex region IDs

| ID | Region | |---|---| | 213 | Moscow | | 2 | Saint Petersburg | | 51 | Samara | | 54 | Yekaterinburg |

Official docs

• API reference: https://www.rush-analytics.ru/api
• Rush Analytics homepage: https://www.rush-analytics.ru

This is an unofficial community wrapper. MIT License. Not affiliated with Rush Analytics.