QualiTaTi MCP server

QualiTaTi for AI agents

Run AI-moderated interviews, conversational surveys and qualitative analysis on QualiTaTi from Claude Code, Codex, OpenCode, Cursor, Claude Desktop or any MCP client.

One hosted MCP server does the work; this repo packages it as an installable plugin for the tools that have plugin systems and documents the one-line config for the rest. The plugin adds three skills that teach the agent QualiTaTi's workflows (setting up an interview study, building a survey, reading results) and two slash commands.

Prerequisites

1. A QualiTaTi account — free tier is fine.
2. A personal API key: qualitati.com → Profile → API keys (keys start with qt_). The key inherits your plan, quotas and billing; revoke it there any time.
3. Put it in your shell: export QUALITATIAPIKEY=qt_…

Clients that implement MCP authorization (OAuth 2.1 with dynamic client registration) can skip the key entirely: point them at the server URL and a QualiTaTi consent page opens in your browser. Claude Code: claude mcp add --transport http qualitati https://api.qualitati.com/mcp, then /mcp → Authenticate.

Install

Claude Code

/plugin marketplace add lanceyuu/qualitati-agent-plugins
/plugin install qualitati@qualitati-agent-plugins

You get the qualitati MCP server (reads QUALITATIAPIKEY), the skills, and /qualitati:new-study and /qualitati:results. Without the plugin, the server alone:

claude mcp add --transport http qualitati https://api.qualitati.com/mcp \
  --header "Authorization: Bearer $QUALITATI_API_KEY"

Codex CLI / app

codex plugin marketplace add lanceyuu/qualitati-agent-plugins
codex plugin add qualitati@qualitati-agent-plugins

Or configure the server directly in ~/.codex/config.toml:

[mcp_servers.qualitati]
url = "https://api.qualitati.com/mcp"
bearer_token_env_var = "QUALITATI_API_KEY"

OpenCode

{
  "mcp": {
    "qualitati": {
      "type": "remote",
      "url": "https://api.qualitati.com/mcp",
      "enabled": true,
      "headers": { "Authorization": "Bearer {env:QUALITATI_API_KEY}" }
    }
  }
}

Cursor

.cursor/mcp.json (or use the Add to Cursor button on https://qualitati.com/developers):

{
  "mcpServers": {
    "qualitati": {
      "url": "https://api.qualitati.com/mcp",
      "headers": { "Authorization": "Bearer qt_your_key_here" }
    }
  }
}

Claude Desktop

Custom connectors cannot send a fixed header yet, so bridge with mcp-remote in claudedesktopconfig.json:

{
  "mcpServers": {
    "qualitati": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://api.qualitati.com/mcp",
               "--header", "Authorization: Bearer qt_your_key_here"]
    }
  }
}

MCP Registry

Listed as io.github.lanceyuu/qualitati on the official MCP Registry (server.json in this repo), so registry-aware clients and directories find it without any of the above.

What the agent can do

A human participant still conducts the interview via the share link; the agent sets studies up and reads the results.

Repo layout

One plugin, two package layouts, shared skills:

Shortened. The full README is on GitHub.