Use the Model Context Protocol (MCP) to run design critiques directly from Claude Desktop, Cursor, or any compatible AI agent.
The Model Context Protocol (MCP) is an open standard that lets AI agents connect to external tools and services. When you add SiteCritic as an MCP server, your AI agent can run deep and flash design critiques and ask follow-up questions — all without leaving your editor or chat.
sitecritic_analyzeAnalyze WebsiteRecord a video walkthrough of a website and generate a structured AI design critique. Use mode ‘deep’ (Gemini Pro, 10 credits) for thorough analysis or ‘flash’ (Gemini Flash, 1 credit) for a faster review. Both return scores, timeline notes, recommendations, and quick wins.
sitecritic_ask_analysisAsk About AnalysisAsk follow-up questions about a completed report’s structured analysis data — scores, recommendations, quick wins, and more.
sitecritic_ask_videoAsk About VideoAsk follow-up questions about a completed report’s video recording with timestamp references. Higher-fidelity visual context.
sitecritic_statusCheck StatusCheck your current credit balance, plan, and billing period. No cost.
Sign in first, then generate a key here.
Paste the config snippet into your agent's MCP settings. Works with Claude Desktop, Cursor, and any MCP-compatible client.
Add to your MCP settings
{
"mcpServers": {
"sitecritic": {
"url": "https://www.sitecritic.ai/api/mcp",
"headers": {
"Authorization": "Bearer sk_live_YOUR_KEY"
}
}
}
}Add to your MCP settings (.cursor/mcp.json)
{
"mcpServers": {
"sitecritic": {
"url": "https://www.sitecritic.ai/api/mcp",
"headers": {
"Authorization": "Bearer sk_live_YOUR_KEY"
}
}
}
}If your client doesn't support remote MCP servers natively, use the mcp-remote proxy:
npx mcp-remote https://www.sitecritic.ai/api/mcp --header "Authorization: Bearer sk_live_YOUR_KEY"
| Plan | Credits | Rate Limit |
|---|---|---|
| Free | 2/month | 10/hour |
| Builder $29/mo | 200/month | 50/hour |
Top up: $10 for 150 credits.