Kameleoon MCP. Manage A/B tests and audience data from your agent.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Kameleoon MCP Server manages A/B testing, personalization, and experimentation workflows directly via your AI client. Use your agent to list active experiments, check variations, or request performance reports without logging into the dashboard.
It lets you define audience segments and query specific targeting rules, making campaign analysis immediate and scriptable.
What your AI agents can do
Create experiment
Creates a new A/B test, requiring a name and a site ID.
Get experiment
Retrieves the full details and configuration for a specific experiment.
Get experiment results
Starts the process to get a performance report for a specified experiment.
Starts a new A/B test by providing a unique name and a specific site ID.
Retrieves specific configurations and metadata for an existing experiment ID.
Initiates an asynchronous process to generate a results report for a completed experiment.
Fetches detailed information for a specific digital property or site.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Kameleoon MCP Server: 10 Tools for Experimentation
Use these tools to programmatically control A/B testing, audience segmentation, and personalization workflows within the Kameleoon platform.
019d75bfcreate experiment
Creates a new A/B test, requiring a name and a site ID.
019d75bfget experiment
Retrieves the full details and configuration for a specific experiment.
019d75bfget experiment results
Starts the process to get a performance report for a specified experiment.
019d75bfget site
Fetches detailed information for a specific site ID.
019d75bflist custom data
Retrieves a list of custom data dimensions used for targeting.
019d75bflist experiments
Lists all active and past experiments within your Kameleoon account.
019d75bflist segments
Lists all defined audience segments and their associated targeting rules.
019d75bflist sites
Lists all digital properties and sites registered to your account.
019d75bflist targeting rules
Lists all active targeting rules governing segment access.
019d75bflist variations
Lists the specific A/B variations tied to a given experiment ID.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Kameleoon, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Your AI client can handle all your A/B testing and personalization workflows right here. You'll use your agent to list all active and past experiments, check specific variations for an experiment ID, and pull performance reports for any test you want to analyze. You've got tools to manage the basics too: you can list all your sites and get detailed info about a specific site ID.
Need to know what's going on with your audience? You can list all defined audience segments and check the targeting rules that govern them or list all active targeting rules. When you're ready to test something new, you can create a new A/B test by giving it a name and a site ID.
You can also pull the full config and metadata for an existing experiment ID. You'll get a list of custom data dimensions used for targeting when you run list_custom_data.
How Kameleoon MCP Works
- 1 Subscribe to the Kameleoon MCP Server and provide your Client ID and Client Secret.
- 2 Your AI client executes a command, like
list_segments, to define the required scope (e.g., 'Show me all segments related to premium users'). - 3 The server executes the API call and returns a structured list of segments, rules, or experiment metadata directly to your agent.
The bottom line is, your agent reads and writes complex testing data without you ever touching the Kameleoon web UI.
Who Is Kameleoon MCP For?
Product Managers who need to monitor live experiment status without logging into the dashboard. Growth Engineers who need to quickly check targeting rules or custom dimensions from their IDE. Data Analysts who need to automate the retrieval of experiment results and metadata for reporting. If you spend time copying and pasting dashboard data, this is for you.
Checks the status of live experiments and confirms variant setups without navigating the Kameleoon dashboard.
Queries specific targeting rules or custom data dimensions directly from their IDE or terminal to validate campaign setup.
Automates fetching experiment results and metadata, piping the data directly into reporting pipelines for analysis.
What Changes When You Connect
- Automate Results Reporting: Don't manually pull data. Use
get_experiment_resultsto request performance reports and feed the raw data directly into your data warehouse. - Validate Targeting Rules: Before launch, run
list_targeting_rulesto confirm the exact parameters governing user segments. This saves time debugging why traffic isn't reaching the right group. - Scope Management: Use
list_segmentsandlist_sitestogether. You can define the audience scope (segments) and ensure the test runs on the correct property (site) in a single agent call. - Deep Debugging: Need to know why a variant isn't performing? Use
list_variationswith an experiment ID to confirm the setup, thenget_experimentto check the overall configuration. - Streamline Setup: The
create_experimenttool lets you launch a new test programmatically, bypassing manual form filling and getting straight to data collection. - Know Your Dimensions: Use
list_custom_datato see exactly what custom data dimensions are available for advanced audience segmentation, ensuring your tests are measurable.
Real-World Use Cases
Checking Campaign Scope Before Launch
A Growth Engineer needs to confirm if the new 'Checkout Flow' experiment will only run on the 'US E-commerce' site. They ask their agent to run list_sites first, confirm the site ID, and then run list_targeting_rules to verify the geographical scope before using create_experiment.
Investigating Low Conversion Rates
A Product Manager notices a segment isn't converting. They ask their agent to run list_segments to see the definition, then list_variations for the related experiment ID, and finally get_experiment to check the overall setup. This pinpoints if the issue is the segment definition or the variant itself.
Auditing All Available Test Parameters
A Data Analyst wants a complete audit of the platform. They run list_experiments to see all active tests, list_custom_data to check available dimensions, and list_segments to map out every possible audience group.
Debugging a Missing Site Connection
A developer needs to know if a new staging site is recognized by Kameleoon. They run list_sites to pull the full inventory. If the site isn't listed, they know the environment needs to be added before they can proceed with create_experiment.
The Tradeoffs
Assuming data is fresh
Running get_experiment_results and getting an error, assuming the API is broken. The issue is often that the results are asynchronous and haven't finished processing yet.
→
Always check the documentation for asynchronous status. If results are pending, your agent needs to poll the endpoint or wait a set time before calling get_experiment_results again.
Mixing up scope and variations
Trying to list variations without knowing the specific experiment ID, or vice versa. The tools are highly dependent on context.
→
First, run list_experiments to get the desired experiment ID. Then, pass that ID directly into list_variations to get the correct list of variations.
Ignoring site context
Running create_experiment without a specific site ID, causing the test to fail or run on an unintended default property.
→
Always run list_sites first. Select the correct, fully qualified site ID, and pass that ID into the create_experiment function call.
When It Fits, When It Doesn't
Use this server if your workflow is inherently data-driven and structured. You need to move beyond viewing dashboards and actually control the testing parameters (segments, variations, experiment status). Use it to define scope (via list_segments and list_targeting_rules), define the test (via list_variations and get_experiment), and measure the outcome (via get_experiment_results).
Don't use this if you just need general marketing advice or if your core problem is generating creative ideas. If you need to compare data from a completely separate source (like a third-party CRM), this server won't help. It only manages the Kameleoon platform. If you just want to read a status, list_experiments is enough, but if you need to act or calculate, you need the full suite of tools.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Kameleoon. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Checking campaign status shouldn't involve three different tabs and copy-pasting IDs.
Today, checking a live experiment's status means jumping between the 'Experiments' tab, the 'Audience' tab to verify segments, and then the 'Results' tab to pull a raw data export. It's a manual, multi-step process that loses context and takes forever.
With this MCP server, you ask your agent to 'List active experiments and their segments.' Your agent calls `list_experiments` and `list_segments` in sequence, compiling the full status—including which segments are attached—and handing you a clean, structured answer. You get the data, not the dashboard.
Kameleoon MCP Server: List Variations MCP Server
Before, finding all the variations for an experiment required manually navigating to the experiment, then clicking a sub-menu, and hoping the list was complete. You had to copy the ID first, and then manually check the variations page.
Now, your agent calls `list_variations` with the experiment ID. It immediately returns a structured list of every variation—Reference, Variant A, etc.—confirming their status and ensuring no variations are missed. It's a single, reliable function call.
Common Questions About Kameleoon MCP
How do I use the `list_segments` tool? +
Use list_segments to retrieve a list of all defined audience segments. This tool shows you the segment names and the specific targeting rules attached to each one, helping you confirm your audience scope.
Is `get_experiment_results` real-time? +
No, get_experiment_results is an asynchronous process. It requests a report, which means the data is generated over time. Your agent needs to be prepared to poll this tool or wait before expecting the final numbers.
What is the difference between `list_experiments` and `get_experiment`? +
list_experiments gives you a high-level overview of all experiments. get_experiment requires a specific experiment ID and returns all the deep metadata, configurations, and details for that single test.
How do I know which site to test on? +
You must use list_sites first. This tool gives you a complete inventory of every site and property registered to your account, ensuring you use the correct environment ID when creating or modifying tests.
How do I use the `list_targeting_rules` tool? +
This tool lists all defined targeting rules. You pass the rule ID to get details, or run it without arguments to see all active rules for the account.
What information can I get using `list_variations`? +
It returns variations associated with a specific experiment ID. You get the variation name, its type, and its current status (e.g., active, paused).
Does `get_site` only show production environments? +
No, get_site retrieves all registered sites. This includes staging, development, and production environments, letting you ensure accurate targeting.
What should I do if `create_experiment` fails? +
Check the error message for common issues like missing site IDs or invalid names. You must provide both a unique name and a valid site ID to successfully create an experiment.
How can I check if an experiment is currently active on a specific site? +
Use the list_experiments tool to get a summary of all experiments, which includes their current status (e.g., 'ACTIVE', 'PAUSED') and assigned site IDs.
Can I see the defined variations for a particular experiment? +
Yes, the list_variations tool takes an experiment ID and returns all variants configured for that test, along with their metadata.
Is it possible to list targeting rules and segments? +
Absolutely. You can use the list_segments and list_targeting_rules tools to inspect how your traffic is being allocated across your experimentation ecosystem.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Password Manager Export Analyzer
Analyze Bitwarden, LastPass, or 1Password CSV exports for weak and duplicate passwords — without EVER sending real passwords to the AI.
Apidog
Sync your Apidog workspace with your AI agent — list projects, fetch endpoint schemas, and export full OpenAPI specifications instantly.
CData Connect Cloud
Universal Data Gateway mapping explicitly proxy structures parsing SQL schemas dynamically connecting APIs natively.
You might also like
Lob
Send physical mail programmatically with an API for postcards, letters, and checks that integrates into your digital workflows.
Placer.ai
Access foot traffic analytics — track visit trends, demographics, and trade areas directly from any AI agent.
Systran (Machine Translation API)
Professional machine translation and linguistic analysis via Systran — translate text, detect languages, and perform dictionary lookups.