# Apache APISIX MCP for AI Agents AI Agent Connect > Apache APISIX MCP lets you manage your cloud-native API gateway using natural language. You can configure routes, set up services, manage upstreams, and control consumers through the APISIX Admin API. It turns complex JSON configuration into a conversation, letting you handle traffic management, load balancing, and rate limiting without manually editing files or using a dashboard. ## Overview - **Category:** loved-by-devs - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_uvuQ6vDEh3EK8RpcSSCqKlnF1NufI3SlkBIczBCN/ai-agent-connect - **Tags:** api-gateway, traffic-management, cloud-native, load-balancing, microservices ## Description This Apache APISIX MCP connects your gateway to any AI agent to orchestrate high-performance traffic management through natural language. Managing a high-traffic gateway usually involves a lot of context switching. You're jumping between your IDE, a web dashboard, and a terminal, constantly copy-pasting JSON snippets to update routes or change load balancing weights. One small typo in a plugin config can take down your traffic, making every change feel risky. This setup changes that by letting you talk to your infrastructure instead of typing out boilerplate. You can ask your agent to spin up a new service, group your upstreams, or swap out SSL certificates just by describing what you need. It handles the heavy lifting of the admin API, so you don't have to worry about the syntax of every single configuration block. If you're looking for a way to keep your microservices organized without the manual overhead, this is the way to do it. It fits right into the Vinkius catalog as a way to turn your gateway into a conversational interface. You get to focus on the logic of your traffic flow while the agent handles the specific execution of creating routes or updating plugin configs. It makes the whole process of managing cloud-native traffic feel like a conversation rather than a series of manual entries. ## Tools ### delete_ssl Deletes a specific APISIX SSL certificate. Use this to remove expired or unused certificates. ### delete_upstream Deletes a specific APISIX Upstream. Use this to remove a backend cluster from your configuration. ### dump_control_discovery Provides a memory dump of discovered service endpoints. This is great for debugging service discovery. ### dump_control_routes Dumps all configured Routes from the Control API. This gives you a full list of your API paths. ### dump_control_services Dumps all configured Services from the Control API. This shows how your backend services are organized. ### get_consumer_group Retrieves a specific APISIX Consumer Group by ID. Use this to check group-level permissions. ### get_consumer Retrieves a specific APISIX Consumer by username. This helps you find a specific user's access level. ### get_control_schema Retrieves JSON schemas for all resources and enabled plugins. Use this to understand the required config structure. ### get_plugin_config Retrieves a specific APISIX Plugin Config by ID. This shows you the specific settings for a plugin. ### get_proto Retrieves a specific APISIX Proto by ID. Use this to check your custom protocol settings. ### get_ssl Retrieves a specific APISIX SSL certificate by ID. This shows you the details of a specific certificate. ### list_consumers Lists all APISIX Consumers. This provides a full list of users with API access. ### list_global_rules Lists all APISIX Global Rules. Use this to see rules that apply across your entire gateway. ### list_plugin_configs Lists all APISIX Plugin Configs. This shows all the active logic like auth or rate limiting. ### list_protos Lists all APISIX Protos. Use this to see your custom protocol definitions. ### list_routes Lists all APISIX Routes. This is the fastest way to see your entire API surface area. ### list_services Lists all APISIX Services. Use this to see how your backends are grouped. ### list_ssls Lists all APISIX SSL certificates. Use this to manage your certificate inventory. ### list_stream_routes Lists all APISIX Stream Routes for L4 routing. Use this to see your TCP/UDP routing setup. ### put_route Creates or updates an APISIX Route. This is how you map URLs to your backend services. ### put_service Creates or updates an APISIX Service. This groups your upstream nodes into logical units. ### put_ssl Creates or updates an APISIX SSL certificate. This manages your gateway's encryption keys. ### put_upstream Creates or updates an APISIX Upstream. This is how you manage your backend clusters. ### trigger_control_gc Triggers a full garbage collection in the HTTP Lua VM. Use this to clear memory during high load. ### delete_stream_route Deletes a specific APISIX Stream Route. This removes L4 routing configurations. ### dump_control_plugin_metadatas Dumps all plugin metadata from the Control API. Use this to audit what plugins are active. ### get_control_resource_healthcheck Retrieves the health status for a specific resource. Use this to check a single cluster's health. ### list_consumer_groups Lists all APISIX Consumer Groups. Use this to see how your users are tiered. ### list_upstreams Lists all configured APISIX Upstreams. This shows you all the backend pools currently managed. ### put_proto Creates or updates an APISIX Proto. Use this to manage custom protocol configurations. ### reload_control_plugins Triggers a hot reload of all plugins. Use this to apply changes without restarting the gateway. ### dump_control_upstreams Lists all configured upstreams from the Control API. This helps you see your entire backend pool at once. ### get_global_rule Retrieves a specific APISIX Global Rule by ID. Use this to check high-level traffic settings. ### get_route Fetches the details of a specific APISIX Route by ID. This is useful for inspecting specific path configurations. ### get_service Retrieves a specific APISIX Service by its ID. Use this to see how your services are grouped. ### get_stream_route Fetches a specific APISIX Stream Route by ID. This is for inspecting L4 routing configurations. ### get_upstream Retrieves a specific APISIX Upstream by ID. This shows you the individual nodes in a cluster. ### delete_consumer_group Removes an existing APISIX Consumer Group. Use this to clean up old access tiers. ### delete_consumer Deletes a specific APISIX Consumer from the system. This removes individual access keys. ### delete_global_rule Removes a specific APISIX Global Rule. Use this to clear gateway-wide settings. ### delete_plugin_config Deletes an APISIX Plugin Configuration. This removes specific logic like rate limiting or auth. ### delete_proto Removes an APISIX Proto from the configuration. This cleans up your custom protocol definitions. ### delete_route Deletes a specific APISIX Route. Use this to remove old API paths from your gateway. ### put_consumer_group Creates or updates an APISIX Consumer Group. This manages how different user tiers are handled. ### put_consumer Creates or updates an APISIX Consumer. Use this to manage API keys and user permissions. ### put_global_rule Creates or updates an APISIX Global Rule. This applies settings to all traffic across the gateway. ### put_plugin_config Creates or updates an APISIX Plugin Configuration. This lets you tweak specific features like auth or logging. ### delete_service Deletes a specific APISIX Service. Use this to remove a backend service group from the gateway. ### get_control_healthcheck Retrieves the health status of all upstream nodes. This is your go-to for checking backend availability. ### put_stream_route Creates or updates an APISIX Stream Route. Use this for managing L4 traffic. ## Prompt Examples **Prompt:** ``` Show me all the active routes in my gateway. ``` **Response:** ``` I've pulled your active routes. You have 4 currently configured: * **auth-api** (ID: 101) * **user-service** (ID: 102) * **payment-gateway** (ID: 103) * **public-assets** (ID: 104) Would you like to see the details for any of these? ``` **Prompt:** ``` Check the health of the backend cluster. ``` **Response:** ``` Checking the upstream health now... | Node | Status | Weight | | :--- | :--- | :--- | | 10.0.0.1:80 | **Healthy** | 50 | | 10.0.0.2:80 | **Healthy** | 50 | All nodes in the 'backend-cluster' are responding correctly. ``` **Prompt:** ``` Create a new rate limit for the mobile app consumers. ``` **Response:** ``` I've updated the mobile-app-consumer. It now has a rate limit of 1000 requests per minute. I used the consumer group settings to ensure this applies to all mobile clients. ``` ## Capabilities ### Create and update API routes You can tell your agent to map incoming requests to specific backend services. ### Group upstream nodes into services Organize your backend clusters into logical service entities for easier maintenance. ### Manage load balancing and health checks Adjust node weights and monitor the health of your backend infrastructure. ### Control consumer access and rate limits Set up per-user limits and authentication rules through simple chat commands. ### Update SSL certificates Refresh and manage your gateway's SSL certificates without manual file handling. ### Trigger hot reloads and garbage collection Refresh plugin configurations or clear memory in the HTTP Lua VM instantly. ## Use Cases ### Rapid Incident Response An SRE notices a spike in 500 errors. They ask the agent to list_upstreams and get_control_healthcheck to identify a failing node in seconds. ### Fast API Deployment A developer needs to deploy a new endpoint. They ask the agent to put_route and put_service to create the new path and backend grouping. ### Security Access Audit A security lead wants to see who has access to the production API. They ask the agent to list_consumers and list_consumer_groups to verify access. ### Traffic Weight Adjustment A team needs to shift traffic to a new cluster. They ask the agent to put_upstream to adjust node weights and balancing algorithms. ## Benefits - Skip the JSON syntax errors by using put_route and put_service to create infrastructure through conversation. - Audit your entire gateway setup instantly using dump_control_routes and dump_control_services to see everything in one place. - Manage consumer access and rate limits easily with put_consumer and put_consumer_group without manual config files. - Monitor your system health in real time using get_control_healthcheck to see which nodes are up or down. - Handle complex plugin configurations through put_plugin_config to avoid manual editing in the dashboard. - Update your security credentials quickly with put_ssl to keep your gateway certificates current. ## How It Works The bottom line is you get a natural language interface for your entire APISIX gateway infrastructure. 1. Subscribe to the Connector and provide your APISIX Admin URL and X-API-KEY. 2. Connect your preferred AI client like Claude or Cursor. 3. Start managing your gateway by asking the agent to list routes or create new services. ## Frequently Asked Questions **What is the Apache APISIX MCP for?** It lets you manage your API gateway using natural language. Instead of writing JSON or using a dashboard, you can tell your AI agent to create routes, manage upstreams, and handle traffic rules. **Can I use this to manage my SSL certificates?** Yes, you can use the Connector to create or update SSL certificates through your AI client. It makes it much easier to keep your gateway secure without manually handling certificate files. **How do I connect my gateway to this Connector?** You just need your APISIX Admin URL and your Admin Key. Once you provide those to your AI client, you can start managing your routes and services immediately. **Does this Connector support load balancing?** Yes, it lets you manage upstreams, which handle your load balancing logic. You can ask your agent to configure node weights, health checks, and balancing algorithms. **Can I use this to troubleshoot my gateway?** Absolutely. You can ask the agent to list routes, check the health of your upstream nodes, or dump the current configuration to see exactly how traffic is flowing. **Is this suitable for production environments?** Yes, it interacts with the official APISIX Admin API. It's a way to perform standard management tasks like route creation and consumer updates through a conversational interface. **Can I update a route's configuration directly through the AI?** Yes. Use the `put_route` tool by providing the Route ID and the new configuration JSON. The agent will interface with the APISIX Admin API to apply the changes immediately. **How do I list all the backend upstreams configured in my gateway?** Simply ask the agent to 'list upstreams'. It will trigger the `list_upstreams` tool and return a detailed list of all virtual hosts and load balancer nodes. **Is it possible to delete a service that is no longer needed?** Yes, you can use the `delete_service` tool. Provide the specific Service ID, and the agent will remove the resource from your APISIX configuration.