# GPTBots MCP > GPTBots lets you manage your entire conversational AI infrastructure directly from your development environment. It gives developers and ops teams direct access to test bot responses, review chat histories, upload knowledge documents, and trigger complex automated workflows—all through a single API connection. ## Overview - **Category:** knowledge-management - **Price:** Free - **Tags:** chatbot, workflow-automation, knowledge-base, conversational-ai, bot-management ## Description You connect this MCP to your agent client, giving it full control over your enterprise AI setup. You can interact with deployed bots by sending messages or listing past conversations. If you need to audit performance, check the chat history of any bot at any time. It’s also how you manage the data powering those bots: list available knowledge documents and upload new content directly to keep the context fresh. Need automation? Trigger configured AI workflows programmatically and even query their execution status. This MCP makes your agent reliable for real-world use, which is exactly what Vinkius delivers by hosting this connection in one place. ## Tools ### list_databases Reads the names of all tables hosted in your platform database. ### create_knowledge_document Allows you to upload new files or create documents within the knowledge base. ### get_conversation Retrieves specific details and the full chat transcript for a single conversation. ### list_conversations Fetches a list of all past chats that occurred with your bots. ### list_knowledge_documents Shows you the titles and metadata of documents currently stored in the knowledge base. ### query_workflow Checks if an automated workflow has finished running, providing its execution status. ### send_bot_message Sends a direct message to one of your deployed AI agents for testing or interaction. ### trigger_workflow Starts an automated, pre-configured workflow sequence immediately. ## Prompt Examples **Prompt:** ``` List all recent conversations for bot ID 'bot_123xyz'. ``` **Response:** ``` Fetching conversations... I found 3 active conversations for this bot. Conversation ID 'conv_987' was updated 5 minutes ago. Would you like to see its chat history? ``` **Prompt:** ``` Trigger the onboarding workflow (ID: 'wf_456') and pass the parameter email='test@example.com'. ``` **Response:** ``` Triggering workflow... Success! The workflow has been initiated. The execution Record ID is 'rec_789abc'. Let me know if you want me to check its status. ``` ## Capabilities ### Manage Bot Conversations View a list of active chats and retrieve the full chat history between a user and an AI agent. ### Update Knowledge Bases Upload new documents or view existing content to keep your bot's contextual information current. ### Orchestrate Workflows Start complex automated processes and check the status of those executions without manually clicking through a dashboard. ### Inspect Data Sources List available tables and records hosted within your platform database for data queries. ## Use Cases ### Auditing Bot Performance After an Incident An Ops team member notices bot performance dropped after a software update. They use list_conversations to pull up chats from the last hour, then call get_conversation on specific IDs to compare chat content before and after the drop, quickly pinpointing where the knowledge context failed. ### Onboarding a New Knowledge Source A Product Manager receives 50 new legal documents. Instead of manually uploading them one by one, they use create_knowledge_document to bulk-upload all files into the knowledge base, ensuring the bot is instantly updated for compliance questions. ### Testing Automated Business Flows A developer needs to test a user onboarding process. They use trigger_workflow to start the process and pass required parameters. Once initiated, they follow up with query_workflow to ensure every step finished successfully before marking it 'live'. ### Building Internal Bot Debugging Tools A developer wants an internal script that verifies agent dependencies. They first list_databases to see all available data tables, then use send_bot_message to test the bot's ability to reference specific records from those tables. ## Benefits - You get instant access to chat history. Instead of logging into a separate dashboard to see what happened last week, you can use list_conversations and then get_conversation to pull up exact transcripts instantly. - Knowledge maintenance is streamlined. You don't have to manually upload files via a web portal; you just call create_knowledge_document from your code to keep the bot current. - Workflow debugging gets easier. If an automated process fails, you can't wait for an email alert. Use trigger_workflow and then query_workflow to check its status right away. - Data visibility is key. By calling list_databases, you immediately see what data sources your bots rely on, which is crucial before writing any code. - Testing agents is faster than ever. Instead of waiting for a manual test run, you can send_bot_message to simulate user input and check the response in real time. ## How It Works The bottom line is you get a single connection point into complex bot management, testing, and data pipelines. 1. Subscribe to this MCP on Vinkius. 2. Provide your GPTBots API Key and Data Center Region credentials. 3. Connect everything via your AI client, then use the tools to manage agents and workflows. ## Frequently Asked Questions **How do I check if my automated workflow ran successfully using GPTBots MCP?** You use the query_workflow tool. First, you call trigger_workflow to start the process, and that action returns a record ID. You then pass this ID into query_workflow to see its current status (running, successful, or failed). **Does GPTBots MCP let me upload new documents for my bot?** Yes. Use the create_knowledge_document tool. This lets you programmatically add new files or update existing knowledge bases to keep your agent's context current. **What is the difference between list_conversations and get_conversation in GPTBots MCP?** list_conversations gives you a high-level summary, showing all recent chats that occurred. You need to use get_conversation, providing a specific conversation ID, if you want to retrieve the full message history. **Can I test my bot's responses without using a web browser?** Absolutely. You can send_bot_message directly from your agent client via this MCP. This lets developers test interactions and see immediate responses right within their IDE. **Is the data I need to query available through list_databases in GPTBots MCP?** The tool simply lists all tables hosted on the platform database. You must then use your agent client's capability to query records within those specific tables.