Enables GitHub Copilot in VS Code to perform UX evaluations with AI personas and human raters.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@autousers Launch a usability test for the new signup flow"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Evaluate UX with AI personas and human raters — directly from Claude, Cursor, ChatGPT, and any MCP-aware client.
The fastest way to connect is via the remote server — no package to install, OAuth handles auth automatically.
Paste https://mcp.autousers.ai/mcp into your client's MCP connector UI or config. OAuth 2.1 launches in the browser on first use.
Client | How to connect |
Claude.ai | Settings → Connectors → Add custom connector → paste URL |
Claude Desktop | Settings → Connectors → Add custom connector → paste URL |
Cursor | See config below or use the one-click deeplink |
VS Code + Copilot | See config below |
ChatGPT | Settings → Connectors → Developer Mode → Add connector → paste URL |
Codex CLI | See config below |
Cursor — ~/.cursor/mcp.json:
{
"mcpServers": {
"autousers": {
"url": "https://mcp.autousers.ai/mcp"
}
}
}VS Code + GitHub Copilot — .vscode/mcp.json (workspace) or ~/.config/Code/User/mcp.json (global):
{
"servers": {
"autousers": {
"type": "http",
"url": "https://mcp.autousers.ai/mcp"
}
}
}Claude Code CLI:
claude mcp add --transport http autousers https://mcp.autousers.ai/mcpCodex CLI:
codex mcp add autousers --url https://mcp.autousers.ai/mcpClients that only support stdio — Cline, Zed, Continue, Goose — use mcp-remote as a shim:
{
"mcpServers": {
"autousers": {
"command": "npx",
"args": ["-y", "mcp-remote@>=0.1.16", "https://mcp.autousers.ai/mcp"]
}
}
}Zed — ~/.config/zed/settings.json:
{
"context_servers": {
"autousers": {
"command": {
"path": "npx",
"args": ["-y", "mcp-remote@>=0.1.16", "https://mcp.autousers.ai/mcp"]
}
}
}
}Continue — ~/.continue/config.yaml:
mcpServers:
- name: autousers
command: npx
args:
- -y
- "mcp-remote@>=0.1.16"
- https://mcp.autousers.ai/mcpFor environments without a browser (CI pipelines, scripted workflows), use the npm package with an API key:
# Claude Code with Bearer token
claude mcp add --transport http autousers https://mcp.autousers.ai/mcp \
--header "Authorization: Bearer $AUTOUSERS_API_KEY"Or run as a local stdio process:
npx -y @autousers/mcpWith environment variables:
{
"mcpServers": {
"autousers": {
"command": "npx",
"args": ["-y", "@autousers/mcp"],
"env": {
"AUTOUSERS_API_KEY": "ak_live_..."
}
}
}
}Related MCP server: UserFlow MCP
39 tools across four categories. Read-only tools carry readOnlyHint=true and are safe to call without side effects.
Tool | Description |
| List question templates available to your team |
| Fetch a single template by ID |
| Create a new team-scoped template |
| Patch a template (only supplied fields change) |
| Hard-delete a template |
| Deep-clone a template into a destination team |
Tool | Description |
| List evaluations the caller can see |
| Fetch one evaluation including config and comparisons |
| Create an SSE or SxS evaluation; optionally queue autouser runs |
| Patch fields on an evaluation |
| Delete an evaluation and its dependent rows |
| Merge wizard fields into a draft evaluation |
| List human + autouser ratings |
| Aggregate stats and per-rater summaries |
| Pairwise Cohen's Kappa inter-rater agreement |
| AI-authored summary, key findings, and recommendations |
| Download results as JSON or CSV |
| Grant a per-user VIEWER / EDITOR / OWNER share |
| List explicit per-user shares |
| Transfer evaluation ownership to another user |
Tool | Description |
| List autousers (built-in + custom) |
| Fetch a single autouser by ID |
| Create a team-scoped custom autouser |
| Patch a custom autouser |
| Soft-delete a custom autouser |
| Deep-clone an autouser into a team |
| Queue autouser runs against an evaluation |
| Cancel pending or running autouser runs |
| Run statuses and summary counts for an evaluation |
| Fetch one autouser run with full context |
| Per-turn token and cost telemetry for a run |
| Compute Cohen's Kappa vs human ratings |
| Get calibration status |
| Freeze a rubric version and set it as active |
| Send disagreements to AI for rubric suggestions |
Tool | Description |
| List every team the caller belongs to |
| Fetch one team's detail |
| List team members with roles and profiles |
| Usage rollup: free-run pool remaining and token spend |
Three readable resource URIs:
URI | Description |
| Full evaluation object including results |
| Template definition with all dimensions |
| Autouser persona with rubric |
Five canned workflows registered as MCP prompts:
Prompt | Description |
| Create an SSE evaluation against a URL and surface results |
| Create an SxS evaluation between two URLs |
| Summarise evaluation results with key findings |
| Run calibration and freeze the rubric when stable |
| Surface autouser/human disagreements and suggest fixes |
When you add the server URL to any supported client, an OAuth 2.1 PKCE + DCR flow launches in the browser. After approving, the client receives a short-lived access token (~15 min) that rotates automatically via a refresh token. No static credentials are stored on the client.
Best for: Claude.ai, Claude Desktop, Cursor, VS Code, ChatGPT — any interactive session.
API keys are long-lived bearer tokens for headless environments. Pass them as Authorization: Bearer ak_live_... — either as an HTTP header in the direct-URL clients or as the AUTOUSERS_API_KEY env var for stdio.
Best for: CI pipelines, Claude Code with --header, Codex CLI --bearer-token, scripts.
Mint keys at app.autousers.ai/settings/api-keys. Keys are shown once at creation — store them in a secrets manager immediately.
Scope | Grants |
| List and fetch templates |
| Create, update, delete, duplicate templates |
| List, fetch, export, view results |
| Create, update, delete, share evaluations |
| List autousers, fetch runs and telemetry |
| Create, update, delete, run, calibrate autousers |
| List ratings for an evaluation |
Variable | Default | Description |
| — | Bearer token for headless / stdio auth ( |
|
| Override the API host (e.g. for self-hosted or local dev) |
The default base URL points to production. You only need AUTOUSERS_BASE_URL if you are targeting a different environment.
Run the bundled doctor command to verify your configuration:
npx -y -p @autousers/mcp autousers-mcp-doctorIt checks that AUTOUSERS_API_KEY is set (for stdio auth), that the API host is reachable, and that at least one tool call succeeds. Exits non-zero on any misconfiguration so it integrates cleanly into CI pre-flight checks.
autousers.ai — product homepage
Install hub — client-specific setup guides
npm package — @autousers/mcp
MCP specification — Model Context Protocol
License — MIT
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Deploy AI user personas to validate user journeys at scale. Find UX friction before real users do.
Run user research from any AI tool. Create studies, recruit participants, query insights.
A flock of AI users tests your deployed app and reports where real people get stuck, with fixes.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/autousers-ai/mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server