• ENVIRONMENT
  • BATCHES
  • SESSIONS
  • HUMAN

How to analyze visual test failures with your AI

Connect Applitools to your AI client to turn messy visual mismatch logs into clear, actionable debugging steps.

Ask AI about this page

Short answer

How can I find the cause of my visual test failures?

Your AI uses the Applitools Connector to scan your test batches and individual sessions. It identifies specific mismatches by checking the browser and OS metadata for every failed result. You get a direct answer on where the visual regression lives.

Where every issue lands.

Debugging outcomes

The AI categorizes your visual failures into these specific buckets.

ENVIRONMENT

Identify OS and Browser

The AI compares failure data across different environments. You'll see if a bug is isolated to Chrome on macOS or widespread across all mobile browsers.

BATCHES

Grouped failure analysis

The AI looks at entire test runs to find patterns. It tells you if a specific deployment caused a spike in mismatches.

SESSIONS

Deep dive into sessions

The AI pulls specific session details to show you exactly what changed. This helps you see the visual difference between the baseline and the failure.

HUMAN

Manual triage required

The AI flags the mismatch and the environment, but you still need to look at the UI to decide if it's a real bug or a false positive.

The workflow

What your AI does when the failures arrive.

The AI handles the heavy lifting of searching through Applitools data so you don't have to.

  1. Locate recent runs

    The AI finds your most recent test executions to start the investigation.

    list_batches

  2. Inspect batch status

    The AI checks the health of a specific run to see how many tests actually failed.

    get_batch

  3. Isolate failed tests

    The AI pulls a list of every failed result within that batch to find the mismatches.

    list_results

  4. Extract environment data

    The AI pulls the specific session details to find the OS and browser responsible for the error.

    get_session

Try it

Copy these to start.

Use these prompts to kick off your investigation.

Starting points

Swap out the batch IDs or session names with your actual Applitools data.

Applitools Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_1ndLMlxAw2nCuJj6kBoTmkGgowFbzClCzzhDWYqC/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Applitools capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "applitools-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_1ndLMlxAw2nCuJj6kBoTmkGgowFbzClCzzhDWYqC/mcp"
    }
  }
}
Copy into chat04

  • List my most recent Applitools batches so I can pick one to analyze.

  • Look at batch ID 12345 and tell me how many visual mismatches occurred.

  • Check the results in batch 67890 and find any failures happening on Safari.

  • Get the details for session ID abc-123 and tell me which OS caused the mismatch.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Applitools once, then ask.

Just link your account via the one-click setup. Your credentials stay encrypted, and you can immediately start asking questions about your test data.

Connect Applitools to your AI

FAQ

How this task behaves

  • 01

    Can the AI fix the visual bugs for me?

    No. The AI identifies the cause and the environment, but you have to fix the code yourself.

  • 02

    Can the AI delete failed test sessions?

    No. The AI only has read access to your Applitools data through these tools.

  • 03

    Does the AI need my Applitools password?

    No. You connect via Vinkius, which manages the connection securely without exposing your password to the AI.

  • 04

    Can the AI see my test screenshots?

    The AI uses the metadata from the sessions to identify mismatches, but it does not directly 'see' the image files.

  • 05

    Can I ask about a specific browser version?

    Yes, as long as that version information is included in the session metadata provided by Applitools.

  • More questions about Applitools? See everything the Applitools Connector can do

Connect Applitools to Claude, Cursor, ChatGPT & more