Connect Claude to Eliciteer via MCP. Claude can create AI interviews, share links with people, and retrieve structured results when interviews complete.
Claude MCP documentationClaude connectors are MCP (Model Context Protocol) servers that extend Claude with tools and data. Eliciteer exposes three tools at https://eliciteer.ai/mcp:
Create an interview from a briefing and get a shareable link.
Check progress, topics covered, and partial results.
Fetch collected information and summary when done.
In Claude.ai, add a custom connector pointing to the Eliciteer MCP endpoint. Remote connectors use Streamable HTTP over HTTPS.
{
"type": "url",
"url": "https://eliciteer.ai/mcp",
"name": "eliciteer",
"authorization_token": "YOUR_MCP_TOKEN"
}Set ELICITEER_MCP_TOKEN on your Eliciteer deployment and use the same value as authorization_token. If the token is not configured, the endpoint accepts unauthenticated requests (not recommended for production).
Add Eliciteer to your Claude Desktop MCP config. Use mcp-remote to bridge stdio clients to the remote HTTP endpoint.
{
"mcpServers": {
"eliciteer": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://eliciteer.ai/mcp"]
}
}
}With authentication:
{
"mcpServers": {
"eliciteer": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://eliciteer.ai/mcp",
"--header",
"Authorization: Bearer YOUR_MCP_TOKEN"
]
}
}
}Any MCP client that supports remote HTTP can connect directly:
{
"mcpServers": {
"eliciteer": {
"url": "https://eliciteer.ai/mcp"
}
}
}Use Eliciteer to ask our product lead what blockers they see for the Q3 launch. Create the interview, share the link, and summarize results when they're done.Deploy Eliciteer, set your MCP token, and add the connector in Claude.