---
title: "FAQ & limits"
section: "MCP & AI assistants"
description: "Answers about the SmartestQA MCP: which assistants work, privacy, read-only use, revoking access, audit, mobile and desktop projects, and every limit."
url: https://docs.smartestqa.ai/en/mcp-faq.html
audience: [customer, admin]
updated: 2026-11-08
keywords: "mcp faq claude cursor privacy conversations data anthropic read-only revoke access audit mobile desktop limits preview 30 minutes"
---
# MCP & AI assistants: FAQ & limits

Answers to common questions about AI assistants in SmartestQA and the limits of the MCP server.

## In short

**On this page you can**

- Get short answers about clients, privacy, access and revoking it.
- Filter the questions for customers or admins.
- Run a five-minute MCP demo for a prospect.
- Look up every MCP limit in one table.

**How to use it**

1. Pick your audience in the filter bar above the questions.
2. Open a question; follow its link for the full procedure.
3. Check Limits when the assistant stops waiting or a preview expires.

## Questions and answers

### Q: Which assistants can I use?

Claude on claude.ai (through [the connector](https://docs.smartestqa.ai/en/mcp-connect.html#claude-ai)), Claude Desktop ([extension](https://docs.smartestqa.ai/en/mcp-connect.html#claude-desktop) or command), Claude Code and Cursor ([command](https://docs.smartestqa.ai/en/mcp-connect.html#claude-code)), and any [other MCP client](https://docs.smartestqa.ai/en/mcp-connect.html#other-client) that can start a local server. See [Choose a way](https://docs.smartestqa.ai/en/mcp-connect.html#choose).

_Audience: customer_

### Q: Connector or extension: which should I take?

On claude.ai, [the connector](https://docs.smartestqa.ai/en/mcp-connect.html#claude-ai): no install, you sign in with your SmartestQA account. On your own computer with Claude Desktop, either works; [the extension](https://docs.smartestqa.ai/en/mcp-connect.html#claude-desktop) uses your [host’s API key](https://docs.smartestqa.ai/en/agent-pool-agents.html#api-key). For Claude Code or Cursor, [the command](https://docs.smartestqa.ai/en/mcp-connect.html#claude-code).

_Audience: customer_

### Q: Does SmartestQA see my conversations?

No. SmartestQA receives only the tool calls the assistant makes, with their arguments, and answers them. The conversation itself stays with your AI provider.

_Audience: customer, admin_

### Q: Which SmartestQA data goes to the AI provider?

The results of the tools the assistant calls: for example project status, requirement names, steps, run logs, defects. They go to the provider of the assistant (for Claude, Anthropic) under that provider’s terms. Use test environments and [test data](https://docs.smartestqa.ai/en/automate-test-data.html). Passwords are never sent: they are [data source keys you fill in SmartestQA](https://docs.smartestqa.ai/en/data-source-secrets.html#where-secrets-belong).

_Audience: admin_

### Q: Can a user reach more through MCP than in the app?

No. Every call runs as that user, with the same [project access](https://docs.smartestqa.ai/en/company-settings-users.html#project-access) and [permissions](https://docs.smartestqa.ai/en/company-settings-roles.html#matrix). Removing records also needs an Admin role and [a click in a dialog](https://docs.smartestqa.ai/en/mcp-safe-changes.html#confirmations). See [Other safeguards](https://docs.smartestqa.ai/en/mcp-safe-changes.html#guards).

_Audience: admin_

### Q: Can I give someone read-only access through MCP?

For a local install, yes: set `SMARTEST_READONLY=true` in the server’s environment (see [Read-only connection](https://docs.smartestqa.ai/en/mcp-connect.html#manage)); the assistant then sees only reading tools. The connector has no read-only switch; its access follows [the user’s role](https://docs.smartestqa.ai/en/company-settings-roles.html#matrix).

_Audience: admin_

### Q: How do I take access away?

Connector: [disconnect SmartestQA](https://docs.smartestqa.ai/en/mcp-connect.html#claude-ai) in Claude under **Settings › Connectors**; a token already issued expires within the hour. Local MCP: its key stops working when its host is [deleted in Test Agents](https://docs.smartestqa.ai/en/agent-pool-agents.html#delete-agent); also remove the extension or run [`smartest-mcp --uninstall`](https://docs.smartestqa.ai/en/mcp-connect.html#manage).

_Audience: admin, customer_

### Q: Can I see what an assistant changed?

Yes. **Configuration** › **Logs** › [**Audit logs**](https://docs.smartestqa.ai/en/logs-audit-log.html) [records every change](https://docs.smartestqa.ai/en/logs-audit-log.html#what-is-recorded) with who made it and from where; [MCP changes](https://docs.smartestqa.ai/en/logs-audit-log.html#mcp-changes) show the source [**MCP (Claude)**](https://docs.smartestqa.ai/en/mcp-safe-changes.html#guards).

_Audience: admin_

### Q: Does it work for mobile and desktop projects?

Yes. The step rules cover [web](https://docs.smartestqa.ai/en/automate-steps-browser.html), [mobile](https://docs.smartestqa.ai/en/automate-steps-mobile.html) and [desktop](https://docs.smartestqa.ai/en/automate-steps-desktop.html) steps. In a [MOBILE project](https://docs.smartestqa.ai/en/quick-start-project.html#project-type) with [Android and iOS](https://docs.smartestqa.ai/en/project-settings-mobile.html), the assistant asks which platform to [dry-run on](https://docs.smartestqa.ai/en/automate-dry-run.html#mobile-dry-run). Desktop and local mobile runs need [your agent online](https://docs.smartestqa.ai/en/agent-pool-agents.html#agent-status).

_Audience: customer_

### Q: Where do the assistant’s dry runs run?

On the SmartestQA server or on your [online agent](https://docs.smartestqa.ai/en/agent-pool-agents.html#agent-status) (see [where dry runs run](https://docs.smartestqa.ai/en/automate-dry-run.html#web-targets)); the assistant asks or follows what you say (“on my agent”). See [Do you need an agent?](https://docs.smartestqa.ai/en/quick-start-connect.html#agent).

_Audience: customer_

### Q: How is this different from Autonomous Agents?

With MCP you work with the assistant in a conversation and approve each change as you go. [Autonomous agents](https://docs.smartestqa.ai/en/ai-agents.html) run alone [on a schedule or after a deploy](https://docs.smartestqa.ai/en/ai-agents-settings.html#trigger) and leave a [report](https://docs.smartestqa.ai/en/ai-agents-settings.html#reports) and [proposals](https://docs.smartestqa.ai/en/ai-agents-runs.html#proposals) for later.

_Audience: customer_

### Q: Do I need a particular Claude plan?

Claude decides which of its plans allow custom connectors and extensions; check your plan in Claude. In a Claude Team or Enterprise organization an owner may have to allow the connector first. On the SmartestQA side, your company’s licence decides what you can use; an Admin sees it in [Company Settings › Billing](https://docs.smartestqa.ai/en/company-settings-billing.html).

_Audience: customer_

## Limits

| Item | Limit |
| --- | --- |
| [Tools](https://docs.smartestqa.ai/en/mcp-tools.html#catalog) · [ready-made prompts](https://docs.smartestqa.ai/en/mcp-tools.html#prompts) | 18 · 4 |
| A [preview](https://docs.smartestqa.ai/en/mcp-safe-changes.html#previews) stays valid | 30 minutes, once, while the MCP keeps running |
| Red dry runs before a scenario is [BLOCKED](https://docs.smartestqa.ai/en/mcp-safe-changes.html#scenario-flow) | 3 |
| Wait for a dry run result per call | 45 s by default, at most 50 s (then the assistant asks again) |
| Wait for a pipeline result per call | at most 50 s |
| [Connector](https://docs.smartestqa.ai/en/mcp-connect.html#claude-ai) access token · renewal | 1 hour · 30 days |
| Connector sign-in attempts | 10 failures in 15 minutes, then a 15-minute pause |
| Node.js for [the command](https://docs.smartestqa.ai/en/mcp-connect.html#claude-code) and [the zip](https://docs.smartestqa.ai/en/mcp-connect.html#manual-zip) | 18 or newer |
| [Claude Desktop extension](https://docs.smartestqa.ai/en/mcp-connect.html#claude-desktop) | Windows and macOS |
| Tool results | in English; the assistant answers in your language |
