Skip to main content

Overview

The Writesonic MCP (Model Context Protocol) server lets you query your GEO analytics data directly from Claude conversations. Ask questions about brand visibility, citations, sentiment, and competitor performance across 10+ AI platforms — all without leaving your AI assistant. Server URL:
https://mcp.writesonic.com/mcp

What You Can Do

  • Track brand visibility across ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, and more
  • Analyze citations — see where and how often AI platforms reference your brand and which pages get cited
  • Monitor sentiment — understand how AI models describe your brand, and how the themes shift over time
  • Compare against competitors — benchmark share of voice, citation share, and visibility scores
  • Find your gaps — surface prompts where competitors appear but you don’t, and the third-party pages behind them
  • Get prioritized actions — pull the Action Center’s recommendations: content to create, pages to optimize, and broken pages AI still cites
  • Explore any prompt — investigate a query’s AI answer and search volume even before you track it
  • Manage your setup in chat — create and edit tracked prompts, topics, tags, competitors, and page portfolios, or kick off a Brand Radar report
  • Slice data by dimensions — market, model, topic, date, website, and more

Connection Methods

PlatformHow to Connect
Claude.aiSettings > Connectors > Add custom connector — https://mcp.writesonic.com/mcp
Claude DesktopAdd to claude_desktop_config.json (see Setup Guide)
Claude Codeclaude mcp add writesonic --transport streamable-http https://mcp.writesonic.com/mcp
ChatGPTSettings > Connectors > enable Developer Mode > Create (see Setup Guide)
Team and Enterprise admins can roll the connector out org-wide — see Organization Setup.

How It Works

  1. Connect — Add the Writesonic MCP server URL to your Claude client
  2. Authenticate — Sign in with your Writesonic account via OAuth when prompted
  3. Query — Ask Claude about your GEO data in natural language

Example Conversation

You: Show me my brand’s visibility score trend over the last 3 months, broken down by AI platform. Claude: Let me pull that data from Writesonic. I’ll query the performance report grouped by model and date… Claude calls get_all_projects, then query_performance_report with dimensions ["model", "date"]

Requirements

  • A Writesonic account with any active paid plan
  • At least one project with tracked prompts configured in your Writesonic dashboard

Available Tools

The MCP server exposes 87 tools spanning data fetching, aggregated reports, prompt analytics, citation analytics, actionable opportunities, and project management. See the Tools Reference for full details. The core tools to know:
ToolDescription
get_all_projectsList all accessible projects (call this first)
get_all_promptsBrowse tracked search queries with visibility scores
get_all_topicsGet topic categories grouping related prompts
get_all_tagsGet custom tag labels for filtering
get_all_citationsGet individual citation URLs
get_all_websitesGet tracked websites (brand + competitors)
get_all_keywordsGet keywords found in AI responses
get_all_modelsGet tracked AI platforms
get_all_marketsGet tracked geographic markets
get_all_domainsGet citing domains (aggregated by domain)
query_performance_reportAggregated visibility and share of voice metrics
query_citation_reportAggregated citation counts and domain analysis
query_sentiment_reportSentiment analysis and keyword occurrence metrics
get_kpi_summarySingle-call KPI dashboard with period-over-period deltas
get_actionables_dashboardSummary of all actionable recommendations
tools_infoGet a ready-made execution plan for a complex question

Next Steps

Setup Guide

Platform-specific setup instructions

Tools Reference

Detailed tool parameters and responses

Use Cases & Examples

Real-world prompts and workflows

GEO API Reference

REST API documentation