• LOGIC
  • AUDIENCE
  • VARIATIONS
  • HUMAN

How to audit experiment configurations with your AI

Connect Kameleoon to your AI client to run instant sanity checks on your testing setup. Once connected, your agent can pull live configuration data to spot mismatches or logic gaps.

Ask AI about this page

Short answer

How can I use AI to audit my Kameleoon experiments?

Your AI pulls your current targeting rules, segments, and variations to compare them against your intended test design. You'll get a clear view of whether your audience segments and rules actually align with your experiment goals.

Audit outcomes

Where your findings land.

The AI categorizes its findings based on what it finds in your Kameleoon account.

LOGIC

Targeting rule conflicts

The AI identifies overlapping or contradictory rules that might prevent a test from running correctly. You get a list of specific rules to review.

AUDIENCE

Segment mismatches

It checks if your audience segments are too broad or too narrow for the intended test. This helps you avoid skewed results.

VARIATIONS

Variation discrepancies

The agent compares your active variations against your plan to ensure all intended versions are present. It flags missing or duplicate variations.

HUMAN

Manual fixes required

The AI identifies the specific areas where your configuration is broken. You'll need to log in to Kameleoon to make the actual changes.

The workflow

What your AI does when the audit request arrives.

The AI moves through your Kameleoon data to build a complete picture of your experiment's state.

  1. Fetch experiment details

    The agent pulls the core configuration and settings for a specific test to understand the baseline setup.

    get_experiment

  2. Inspect targeting logic

    It pulls all active rules to see exactly which users are being funneled into the test.

    list_targeting_rules

  3. Verify audience segments

    The AI checks the definitions of your segments to ensure they match your testing criteria.

    list_segments

  4. Validate test versions

    It lists all available variations to confirm that the experiment has the right number of versions active.

    list_variations

Try it

Copy these to start.

Use these prompts to kick off an audit.

Starting points

These are starting points. Replace the specific experiment names or segment IDs with your actual Kameleoon identifiers.

Kameleoon Connector

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

Connector linkhttps://edge.vinkius.com/vk_preview_6gJb8h8nvE83qpO01DFnHsxpfAmKyDZOllD6tBbY/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 — Kameleoon capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "kameleoon-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_6gJb8h8nvE83qpO01DFnHsxpfAmKyDZOllD6tBbY/mcp"
    }
  }
}
Copy into chat04

  • Audit the targeting rules for my 'Summer Sale' experiment to see if they conflict.

  • Check if the 'VIP Customers' segment is correctly applied to my current test.

  • List all variations for the 'Homepage Redesign' experiment and tell me if any look incomplete.

  • Compare my experiment targeting rules against the 'Mobile Users' segment to find mismatches.

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

Start here

Connect Kameleoon once, then ask.

Just link your account via Vinkius. Your credentials stay encrypted, and you can immediately start asking your agent about your test configurations.

Connect Kameleoon to your AI

FAQ

How the audit behaves

  • 01

    Can the AI change my experiment settings?

    No. The AI can only read your configuration data. You must make any changes manually within the Kameleoon interface.

  • 02

    Can the AI delete a segment or a variation?

    No. The tools only allow the AI to list and get data. It cannot delete or modify anything in your account.

  • 03

    How does the AI know if a rule is wrong?

    The AI compares the data it pulls from your targeting rules and segments against the specific logic you describe in your prompt.

  • 04

    Does the AI see all my experiments?

    The AI can see any experiment you ask it to look at, provided you give it the correct identifier or ask it to list them.

  • 05

    Can I audit multiple experiments at once?

    Yes, you can ask your agent to compare settings across different experiments by providing their names or IDs.

  • More questions about Kameleoon? The Connector page answers them. See everything the Kameleoon Connector can do

Connect Kameleoon to Claude, Cursor, ChatGPT & more