# Chaindesk MCP for AI Agents AI Agent Connect > Chaindesk lets you build custom AI agents trained on your own data. It connects your Chaindesk.ai account to your AI client so you can manage knowledge bases, handle customer support, and qualify leads using your proprietary information. ## Overview - **Category:** knowledge-management - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_T5VjYPY8ezoZU63DLLaawQj8Wz1PrnywxCHz077q/ai-agent-connect - **Tags:** llm-training, custom-chatbots, knowledge-retrieval, no-code-ai, rag-pipeline ## Description Chaindesk lets you connect your Chaindesk.ai account to any AI client so you can manage your custom LLM orchestration through natural conversation. Instead of manually uploading files or copy-pasting text to train a bot, you can manage your entire knowledge base from your chat interface. You can build out a fleet of specialized agents, each tied to specific data sources like your company website, internal documents, or text files. It handles the heavy lifting of grounding your AI in your own proprietary information. This means when a user asks a question, your agent pulls from the right place without you having to do any manual retrieval. You can also see how your agents are performing by looking at their chat histories and monitoring the status of your datastores in real time. It turns your AI client into a command center for your company's knowledge. This is one of the many tools you can find in the Vinkius catalog to bridge the gap between raw data and useful agent actions. You get to act as a knowledge architect without the headache of manual data management, making it easier to deploy bots that actually know your business. You can quickly pivot your agents to focus on new products or updated policies just by updating the source data. ## Tools ### create_agent Build a new AI agent by providing a name, datastore, and prompt. This starts your custom bot workflow. ### get_agent Retrieve the details of a specific agent. This lets you check current configurations without leaving your chat. ### get_datastore See the specific details for a single datastore. This is useful for checking which URLs or files are connected. ### get_messages Pull the message history for a specific conversation. Use this to audit what your bot said to a user. ### query_agent Send a message to a specific agent. This is the primary way to get answers from your custom-trained bots. ### update_agent Modify an existing agent's settings. Use this to change system prompts or switch models quickly. ### list_agents See all the AI agents you've built in your account. This helps you manage your bot fleet from one place. ### list_conversations Browse your chat history. You can filter these by a specific agent to find old threads. ### list_datastores View your knowledge collections. This shows you what data your agents currently have access to. ### delete_agent Remove an agent from your account. Use this to clean up old bots or clear out test instances. ### upsert_datasource Add or update a data source like a website URL or a document. This keeps your knowledge base fresh. ## Prompt Examples **Prompt:** ``` List all my available AI agents in Chaindesk. ``` **Response:** ``` I've retrieved your AI agents. You currently have 3 active bots: **Support Bot** (ID: agent_1), **Legal Analyst**, and **Product Research**. Which one would you like to query or update? ``` **Prompt:** ``` Ask my 'Support Bot' (ID: 'agent_1'): 'How do I reset my password?'. ``` **Response:** ``` Querying agent... **Support Bot response**: 'To reset your password, go to the login page and click Forgot Password. You will receive an email with instructions.' I've documented the session ID for context. ``` **Prompt:** ``` Add 'https://vinkius.com/faq' to datastore 'ds_123'. ``` **Response:** ``` Knowledge ingestion triggered! I've added the FAQ URL to datastore ds_123. Chaindesk is now processing the content to update your agents' knowledge base. Need any other sources added? ``` ## Capabilities ### Create multiple AI agents Build and manage several different agent instances with unique prompts and data sources. ### Upload website URLs and documents Add new data sources to your knowledge base by providing links or files. ### Query your proprietary data Get context-aware answers from your agents based on your specific company information. ### View entire conversation histories Access full chat threads to see how your agents are interacting with users. ### Monitor your knowledge collections Check the status of your datastores to ensure your information is being indexed correctly. ## Use Cases ### Automating customer support for a retail site A support lead uses query_agent to see how a bot handles a refund request based on the company's specific policy documents. ### Updating internal company wikis An ops person uses upsert_datasource to add a new product manual, ensuring all internal agents have the latest specs. ### Auditing bot performance for quality assurance A manager uses list_conversations to check for common customer complaints and identify where the bot needs more training. ### Launching specialized research assistants A product lead uses create_agent to launch a new legal research bot tied to a specific datastore of internal contracts. ## Benefits - Build a fleet of bots using create_agent without ever touching a dashboard. - Keep your knowledge base current by using upsert_datasource to add new URLs or files. - Audit every interaction by pulling logs with get_messages for quality control. - Monitor your data health using list_datastores to see what is being indexed in real time. - Update agent instructions on the fly with update_agent to fix behavior issues instantly. ## How It Works The bottom line is you get a direct line from your AI client to your custom-trained knowledge base. 1. Subscribe to the Chaindesk MCP and get your API key from your Chaindesk dashboard. 2. Connect your AI client to the Connector using your unique credentials. 3. Start querying your agents or updating your knowledge base using natural language. ## Frequently Asked Questions **Can I use Chaindesk MCP to train my bots on my own website?** Yes, you can. By adding your website URL as a data source, the Connector allows your AI agent to pull information directly from your site to answer questions. **How do I see what my agents are saying to customers?** You can view entire conversation histories directly. This allows you to audit responses and monitor how your bots are interacting with users in real time. **Can I create multiple agents for different departments?** Yes, you can build a fleet of specialized agents. Each one can have its own unique prompt and its own specific datastore for different business units. **Does Chaindesk MCP help with lead qualification?** It does. By grounding your agents in your specific product data and FAQs, they can accurately qualify leads and provide detailed information to potential customers. **How do I update the information my agent uses?** You can update your knowledge base by adding or modifying data sources. The Connector handles the ingestion so your agents always have the latest information. **Can I monitor my knowledge base status?** Yes, you can monitor your datastores to see which sources are connected and check their current status to ensure your agents are grounded correctly. **How do I find my Chaindesk API Key?** Log in to your Chaindesk.ai account, navigate to **Settings** > **API Keys**, and generate a new key for your integration. **What is a Datastore?** A Datastore is a collection of documents and URLs that your AI agent uses as its knowledge base to answer queries accurately. **Can I maintain conversation context via AI?** Yes! Provide a unique `conversationId` to the `query_agent` tool to maintain historical context across multiple turns with your custom bot.