Guides

MCP Server

Give your AI agents direct access to case study search, company analysis, and vendor comparison.

What is MCP?

Model Context Protocol (MCP) lets AI agents use external tools. The casestudies.dev MCP server gives any MCP-compatible agent (Claude Desktop, Claude Code, ChatGPT with plugins, custom agents) access to our full API through natural conversation.

Setup

Claude Desktop

Add to your claude_desktop_config.json:

1{
2  "mcpServers": {
3    "casestudies": {
4      "command": "python",
5      "args": ["/path/to/mcp-server/server.py"],
6      "env": {
7        "CASESTUDIES_API_KEY": "cs_live_your_key_here"
8      }
9    }
10  }
11}

Claude Code

$claude mcp add casestudies python /path/to/mcp-server/server.py -e CASESTUDIES_API_KEY=cs_live_xxx

Requirements

  • Python 3.10+
  • httpx (pip install httpx)
  • A casestudies.dev API key

Available tools

ToolWhat it does
search_case_studiesSemantic search with natural language queries
get_case_studyGet full details of a specific case study
search_companiesFind companies by name or industry
get_company_analysisCompetitor analysis — top industries, clients, outcomes
compare_vendorsSide-by-side comparison of up to 5 companies
get_company_timelineChronological view for due diligence

Example prompts

Once connected, just ask:

  • “Find case studies about companies that improved onboarding conversion”
  • “What industries does Checkout.com serve? Show me their notable clients”
  • “Compare GoCardless, Checkout.com, and Paddle for ecommerce payments”
  • “Show me how fintech companies have reduced customer churn”

Source code

The MCP server is open source: github.com/casestudies-dev/casestudies/mcp-server