# ConfigCat MCP for AI Agents MCP > ConfigCat manages feature flags and remote configurations directly through your AI client. This MCP lets you list environments, create settings, toggle features, and update configuration values without touching code. It provides granular control over application logic in development, staging, and production. ## Overview - **Category:** ship-it - **Price:** Free - **Tags:** feature-flags, remote-config, release-management, toggles, configuration-management ## Description You shouldn't need to redeploy code every time a feature flag needs tweaking or an environment variable changes. ConfigCat gives developers and product teams the ability to manage all that complexity from chat alone. Using your AI client, you can list environments (like Test, Staging, Production) and instantly check specific setting values for any given configuration. Need to turn on a beta feature? Your agent handles it: you call the necessary tools to create or update a setting value immediately. You can also define user segments, which is critical for running targeted A/B tests or canary releases. This capability means your release workflow moves at the speed of conversation, not the speed of CI/CD pipelines. When you connect ConfigCat via Vinkius, your AI client gets access to this entire catalog of feature toggles and environment controls. ## Tools ### create_config Creates a brand new configuration container for your product. ### create_environment Builds a dedicated environment (like QA or Production) to isolate testing. ### create_segment Defines a new group of users for targeted testing or release management. ### create_setting Adds a specific feature flag or setting (e.g., 'beta_checkout') to an existing configuration. ### delete_config Removes an entire product configuration container. ### delete_environment Deletes a specific testing or live environment. ### delete_segment Removes a predefined user segment from the system. ### delete_setting Deletes an existing feature flag or setting key. ### get_config Retrieves all details about a specific configuration container. ### get_environment Gets detailed information on a specified environment instance. ### get_segment Fetches the criteria and details of a specific user segment. ### get_setting Retrieves metadata about a feature flag or setting key. ### get_setting_value Checks the current value of a setting within a specific environment. ### list_configs Lists all configuration containers available for your product ID. ### list_environments Shows all existing environments (e.g., Test, Staging) tied to a product. ### list_segments Lists all defined user segments for the entire product. ### list_settings Shows every feature flag or setting key within a given configuration. ### update_setting_value Changes the value of an existing setting, which triggers real-time behavior updates in your application. ## Prompt Examples **Prompt:** ``` What are the different environments I have for my main application? ``` **Response:** ``` **Product ID: prod_123** We currently maintain 4 active environments: * 🟢 **Production**: Live customer traffic. (ID: env_prod) * 🟡 **Staging**: Final QA testing before release. (ID: env_staging) * 🟠 **Beta**: Limited internal testing group. (ID: env_beta) * ⚫ **Development**: Local developer sandboxes. (ID: env_dev) ``` **Prompt:** ``` Can you check the value of 'new_checkout_ui' in Staging? ``` **Response:** ``` Found it. The feature flag `new_checkout_ui` is active for the **Staging** environment (`env_staging`). * **Key:** new_checkout_ui * **Type:** Boolean * **Current Value:** True * **Last Updated:** 2026-06-12 (by Product Manager) This means the staging environment is currently running the new UI. ``` **Prompt:** ``` I need to turn off the beta feature flag for everyone. ``` **Response:** ``` Got it. I am updating `beta_feature` in **Production** now, setting its value to `False`. ✅ **Update Successful:** Feature Flag 'beta_feature' set to False. * **Product ID:** prod_123 * **Environment:** Production * **Change Effective:** Immediately This change will be reflected in the application within minutes. ``` ## Capabilities ### View and organize system configurations List all available product configurations, environments, and user segments to understand the current setup. ### Manage core settings and flags Create new feature flags or settings—whether they're boolean switches, strings, or numbers—and delete unused ones. ### Dynamically adjust values in real-time Retrieve the current value of any setting for a specific environment and update that value instantly to trigger immediate application changes. ### Build user targeting groups Define new user segments or view existing ones, allowing you to target feature rollouts to specific user groups. ## Use Cases ### Testing a new payment flow for premium users A product manager needs to test a paid feature only on paying customers. They use the agent to `create_segment` specifically for 'Premium Users' and then verify that the associated setting value is active for that group using `get_setting_value`. ### Rolling back a broken checkout button A developer realizes a new feature flag broke the site. Instead of deploying an emergency hotfix, they ask their agent to use `update_setting_value` to immediately set the problematic flag's value back to 'false'. ### Setting up isolated staging testing A DevOps engineer needs a clean test slate. They instruct their agent to `create_environment` for 'Pre-Prod QA', ensuring it has all necessary settings before the next major build. ## Benefits - Control feature rollouts without code changes. Use `update_setting_value` to flip a switch, activating or deactivating features in production instantly. - Maintain environment parity across your product line. List all configurations using `list_configs` and verify that Staging matches Production before launch. - Target specific users for testing. Create new segments via `create_segment` so you can run A/B tests only on a small, controlled group of users. - Speed up debugging by verifying settings. You can check the exact state of any flag in any environment using `get_setting_value`, eliminating guesswork. - Streamline your release process. By managing environments and configurations with this MCP, you reduce reliance on manual deployment steps. ## How It Works The bottom line is: you talk to your agent, and it executes complex configuration changes across development environments for you. 1. First, subscribe to the MCP and provide your ConfigCat API Key ID and Secret. 2. Second, reference the desired action with your agent—for example, 'What's the current value of the Beta Feature flag in Staging?' 3. Third, your AI client calls the appropriate tools. It returns a clear confirmation or the requested configuration data to you. ## Frequently Asked Questions **How does the ConfigCat MCP help manage feature flags across different environments?** The ConfigCat MCP lets you view and manipulate settings for multiple environments (Test, Staging, Production) from a single chat interface. You can check if the flag status is consistent everywhere, saving hours of manual cross-checking. **Can I use this with my AI client to run A/B tests?** Yes. By managing user segments and dynamically updating setting values, you can easily target a small group (a segment) for an A/B test without affecting the general user base. **What if I need to delete a whole set of flags or settings?** You have tools available to remove configuration containers, environments, and individual settings. This helps keep your product's configuration clean and prevents clutter from old features. **Does the ConfigCat MCP require me to write code for simple flag changes?** No. The entire point of this connector is to manage these controls through natural conversation with your AI client, eliminating the need to commit code just to flip a switch. **Is ConfigCat MCP safe to use in production? How do I know it's secure?** The tool allows you to get and update values for Production environments. Because all actions are logged and initiated through your AI client, you retain an audit trail of every change.