AI test automation, straight from your editor
TestFlair speaks MCP (Model Context Protocol), so the AI assistant in Claude Code, Cursor, VS Code or Claude Desktop can drive the platform for you. Ask it to turn a requirement into user stories, test cases and runnable scripts, run them in the cloud and file the bugs it finds, all against your real projects and with your own permissions.
Illustration: an AI assistant in the editor running the TestFlair pipeline, from picking a workspace to a cloud test run with video and a proposed bug report.
One sentence, the whole pipeline
Using TestFlair, create a project for our checkout flow against our staging site, generate user stories from this requirement, then test cases and a script for the happy path, and run it.”
Your assistant works through the same steps as the web app:
- Step 1
Pick a workspace
Your assistant finds the workspaces you belong to.
- Step 2
Create a project
Name it, point it at your app and give it the requirement.
- Step 3
Generate user stories
From the requirement text or attached documents.
- Step 4
Generate test cases
Step-by-step cases for each user story.
- Step 5
Generate & run scripts
Java automation on Selenium or Playwright, run in the cloud with a video of the session.
- Step 6
File the bugs
Failed runs are analysed into proposed bug reports you review before filing.
Connect in under a minute
Add the server at https://mcp.testflair.ai/mcp (streamable HTTP). The first time your editor connects, it opens TestFlair sign-in in your browser. There's no API key to create or paste.
Set up TestFlair MCP in Claude Code
Run once in your terminal.claude mcp add --transport http testflair https://mcp.testflair.ai/mcpSet up TestFlair MCP in Cursor
Add to .cursor/mcp.json in your project, or your global Cursor MCP settings.{
"mcpServers": {
"testflair": {
"type": "http",
"url": "https://mcp.testflair.ai/mcp"
}
}
}Set up TestFlair MCP in VS Code
Add to .vscode/mcp.json.{
"servers": {
"testflair": {
"type": "http",
"url": "https://mcp.testflair.ai/mcp"
}
}
}Set up TestFlair MCP in Claude Desktop
Add TestFlair as a custom connector and enter this address as its URL.https://mcp.testflair.ai/mcpWhat your assistant can do
More than thirty tools covering the whole testing workflow. Your assistant can also read and correct what's already there by hand. AI generation uses Flairs, just as in the web app.
- 01
Projects & requirements
Create and correct projects, store requirement text and attach requirement documents.
- 02
Test data
Add real input values, and update one entry everywhere the existing test steps use it.
- 03
User stories
Generate them, import them from a document as written, or write and edit them by hand.
- 04
Test cases
Generate cases per story, read them step by step, and correct them in place.
- 05
Scripts & execution
Generate runnable scripts, fix a selector without regenerating, and run them under your execution profiles.
- 06
Bugs
Turn a failed run into proposed bug reports with evidence, file the ones you keep, and update their status.
The full tool reference, with tips and troubleshooting, is in the MCP setup guide.
Frequently asked questions
What is the TestFlair MCP server?
The TestFlair MCP server lets the AI assistant in your editor use TestFlair through the Model Context Protocol (MCP). Add https://mcp.testflair.ai/mcp to Claude Code, Cursor, VS Code or Claude Desktop, sign in once in your browser, and your assistant can create projects, generate user stories, test cases and Java test scripts, run them in the cloud and file the bugs they find — with your own TestFlair account and permissions.
Which editors and assistants does the TestFlair MCP server work with?
Any MCP client that supports streamable HTTP transport, including Claude Code, Cursor, VS Code and Claude Desktop. If your client asks for a transport type, choose http, not sse.
Do I need an API key?
No. The first time your editor connects, it opens the TestFlair sign-in page in your browser. You sign in with your TestFlair account, and the tools become available to your assistant.
Does using TestFlair through MCP cost Flairs?
AI work does, exactly as in the web app: generating user stories, test cases, scripts and bug suggestions, and importing user stories from a document. Running tests does not use Flairs, and neither does reading, listing or editing your projects, stories, test cases, scripts and bugs by hand.
What can my assistant see and change?
Exactly what you can in the web app. Every call runs with your own TestFlair account and project permissions, against your real data.
Can I test an app running on localhost?
Tests run in the TestFlair cloud, not on your machine, so the app must be reachable from the internet. For a local or private app, set up a secure tunnel as described in the Test Execution guide.
into your editor
