# Ada MCP for AI Agents AI Agent Connect > Ada MCP connects your AI agent to your Ada account to manage customer service workflows. It lets your agent handle conversation oversight, sync user metadata, and update your knowledge base in real-time. It's built for teams that need to bridge the gap between their support bot and their internal data. ## Overview - **Category:** customer-support - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_AzLRmp3uK52IN3iowI4nfLlOj7D2qBhC9fZYSMTS/ai-agent-connect - **Tags:** conversational-ai, customer-service-automation, chatbot, knowledge-base, support-automation, user-metadata ## Description When you're running a high-volume support operation, your AI bot is only as good as the data it can access. You end up jumping between tabs to check what a customer actually said, looking up their history in one place, and then manually updating help articles in another. This Connector changes that by giving your AI agent direct access to your Ada account. Instead of you doing the heavy lifting, you can just ask your agent to find specific customer interactions or pull up a user's profile to see their custom attributes. If you notice your bot is struggling with a specific question, you can tell your agent to write a new help article right then and there to fix the knowledge gap. It makes the whole support ecosystem feel like one cohesive unit rather than a bunch of disconnected tools. It's a great way to keep your knowledge base fresh without the usual friction of manual entry. Because Vinkius makes it so easy to connect these types of tools, you can get your agent up and running in minutes. You stop being the middleman for data and start actually managing the strategy of your customer experience. You can oversee the health of your automated systems, identify where your bot is failing, and make real-time adjustments to your content. This connection ensures that your support team spends less time on admin tasks and more time on high-value customer interactions. ## Tools ### get_end_user Get a specific user's profile and their custom metavariables. It's the fastest way to see who you're talking to. ### list_articles See every help article in your knowledge base. Your agent uses these to provide accurate answers to customers. ### create_article Write and publish a new help article directly. This lets you fix bot knowledge gaps in seconds. ### list_conversations Pull a list of past and current support chats handled by the Ada bot. This helps you spot recurring issues quickly. ## Prompt Examples **Prompt:** ``` Show me the last 5 conversations handled by Ada. ``` **Response:** ``` I've pulled the last 5 transcripts for you. Here is a summary: * **Conversation 1**: Resolved (Bot handled refund query) * **Conversation 2**: Resolved (Bot handled login issue) * **Conversation 3**: Handoff (User asked for human agent regarding billing) * **Conversation 4**: Resolved (Bot handled shipping update) * **Conversation 5**: Resolved (Bot handled password reset) Would you like me to summarize the themes from the ones that required a handoff? ``` **Prompt:** ``` What's in our knowledge base about the new refund policy? ``` **Response:** ``` I've checked the help articles. Here is the current policy: **Refund Policy Summary:** * **Eligibility**: Items must be returned within 30 days. * **Condition**: Items must be in original packaging. * **Processing**: Refunds are issued within 5-7 business days. It looks like we have a clear article on this. Would you like me to find the specific article link? ``` **Prompt:** ``` Get the profile for user 99821. ``` **Response:** ``` I've retrieved the profile for User 99821: * **Status**: Premium Subscriber * **Tenure**: 2 years * **Preferred Channel**: Email * **Custom Metavariables**: * `loyalty_tier`: Gold * `account_manager`: Sarah J. * `last_purchase_value`: $450.00 ``` ## Capabilities ### Pull active and past support transcripts Retrieve conversation history to identify trends and specific customer issues. ### Fetch specific customer profiles Access user profiles and custom metavariables for better context during interactions. ### Browse your entire help article catalog Let your agent see every article in your knowledge base to provide accurate answers. ### Add new articles to your knowledge base Create new help content on the fly to improve your bot's response accuracy. ### Identify trends in automated resolutions Get insights on how often your bot resolves issues without human intervention. ### Handle data privacy and retention requests Manage privacy requests and conversation history directly from your chat interface. ## Use Cases ### Fixing bot errors on the fly A manager asks the agent to see the last 5 conversations, notices a mistake, and tells the agent to `create_article` to fix the knowledge gap immediately. ### Contextualizing human handoffs An agent uses `get_end_user` to pull specific metavariables into a summary so a human rep knows exactly who they are talking to before joining. ### Auditing knowledge gaps A training lead asks the agent to `list_articles` to see what's missing and then adds a new one to cover a new product feature. ### Identifying recurring issues A CX manager asks for a list of conversations to find common themes for the next sprint, allowing for data-driven product decisions. ## Benefits - Stop manual tab-switching by letting your agent pull conversation transcripts using `list_conversations`. - Get instant customer context by retrieving profiles and metavariables with `get_end_user`. - Keep your help center current by having your agent write new articles with `create_article`. - Reduce training time by letting your agent browse the existing catalog via `list_articles`. - Improve resolution rates by identifying trends in your support history. - Simplify compliance by managing data privacy requests through a single chat interface. ## How It Works The bottom line is you get a natural language interface for your entire Ada support backend. 1. Subscribe to the Ada MCP on Vinkius. 2. Input your Ada Platform Token and Handle in your AI client. 3. Start asking your agent to manage conversations or update articles. ## Frequently Asked Questions **How does the Ada MCP help with customer support?** It connects your Ada account to your AI agent, allowing it to pull conversation history, check user profiles, and manage your knowledge base using natural language. **Can I use the Ada MCP to update my help articles?** Yes. You can ask your AI agent to create or update help articles directly in your Ada knowledge base without leaving your chat interface. **Does the Ada MCP support user metadata?** Yes, it can retrieve specific user profiles and custom metavariables, giving your agent the context it needs to provide better answers. **How do I connect the Ada MCP to my AI agent?** You can subscribe to the Ada MCP on Vinkius and then enter your Ada Platform Token and Handle into your preferred AI client like Claude or Cursor. **Can the Ada MCP see my past conversations?** Yes, it can list and retrieve both active and past customer support conversations handled by your Ada bot to help identify trends. **Can I use the Ada MCP to manage privacy requests?** Yes, it allows you to manage data privacy requests and conversation retention directly through your AI agent's interface. **Can I update my knowledge base via the agent?** Yes! Use the `create_article` tool to add new information to your Ada knowledge base. This helps the AI bot stay up-to-date with your latest product or policy changes. **What are the API rate limits on Ada?** The default platform API limit is 10,000 requests per day (10 per second). Some specific endpoints, such as the Data Export API, have additional limits like 15,000 requests per day. **Are there quotas per knowledge article in Ada?** Yes, Knowledge Base APIs enforce a constraint of 100KB per individual article, up to a maximum limit of 50,000 global articles in your organization.