# DocsBot MCP for AI Agents AI Agent Connect > DocsBot lets you manage your AI knowledge base and query your custom bots directly from your AI client. It handles bot oversight, source tracking, and interaction logging so you can keep your documentation accurate and your team informed without switching tabs. ## Overview - **Category:** customer-support - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_DUvS3dfoYdN9s8Yd26fUzX2DZZJItzvIEMs195YL/ai-agent-connect - **Tags:** ai-chatbot, knowledge-base, indexing, natural-language-query, documentation-search, ai-training ## Description DocsBot lets you manage your AI knowledge base and query your custom bots directly from your AI client. You can stop guessing if your documentation actually made it into the bot's brain. Instead, you can ask your agent to check the status of a specific PDF or see which URLs are currently being used as sources. If a customer asks a weird question, you can pull the exact interaction log to see how the bot handled it without digging through a messy dashboard. It's about having a direct line to your knowledge base. You can list all the bots your team has set up, check which ones were updated most recently, and even query them in real-time to see if they're giving the right answers. It's a huge time saver for anyone who needs to keep their AI content fresh and accurate. You'll find this is a lot easier to manage when it's all sitting in one place in your Vinkius catalog. ## Tools ### get_docsbot_account_metadata Pulls the basic info for your current account. It's how the agent knows who it's talking to. ### ask_bot_question Sends a question to a specific bot and gets an AI response. This lets you test your bot's knowledge instantly. ### get_bot_knowledge_summary Shows how big your knowledge base is and how many sources it has. It's a quick way to see the scale of your data. ### get_bot_details Grabs the specific settings and metadata for a single bot. Use this to check configuration details on the fly. ### list_recently_indexed_bots Shows which bots were updated lately. It helps you verify that your recent uploads actually went live. ### list_bot_interaction_logs Pulls a list of recent questions and answers from a bot. This is perfect for auditing bot behavior. ### list_team_bots Lists every bot your team has created. It gives you a bird's-eye view of your entire bot fleet. ### list_bot_knowledge_sources Shows every URL, PDF, and document a bot uses. You can use this to see exactly what data a bot is reading. ### list_docsbot_teams Lists the teams you belong to in the platform. It helps you navigate different organizational units. ### search_bot_sources Lets you find specific knowledge files by keyword. It's great for finding that one specific PDF in a sea of docs. ## Prompt Examples **Prompt:** ``` Did our 'Support Bot' actually index the new 2024 handbook yet? ``` **Response:** ``` I checked the recent updates and the **Support Bot** was indexed 2 hours ago. It's currently using 14 sources, including the new handbook. **Status:** - **Index Status:** Live - **Last Updated:** 2024-07-10 14:00 Would you like me to list the specific sources it's using? ``` **Prompt:** ``` Show me some of the recent questions people are asking the Sales Bot. ``` **Response:** ``` Here are the last 5 questions the **Sales Bot** handled: | Question | Status | | :--- | :--- | | What are your pricing plans? | Answered | | Do you offer an annual discount? | Answered | | How do I upgrade? | Answered | | Is there a free trial? | Answered | | Can I pay by check? | Answered | It seems people are asking a lot about annual discounts. Should I pull the full interaction logs for these? ``` **Prompt:** ``` What's the total size of our Tech Docs knowledge base? ``` **Response:** ``` The **Tech Docs** bot currently has a knowledge base summary showing **1,240 total sources**. It's pulling from 12 different team folders. **Source Breakdown:** - **PDFs:** 450 - **URLs:** 790 Do you want a list of all the bots in this team? ``` ## Capabilities ### Query custom bots Ask your specific bots questions and get AI-generated answers based on your documentation. ### Track knowledge sources See exactly which URLs and PDFs are being used to train your bots. ### Monitor indexing status Check which bots have been updated recently and verify your data is live. ### Audit interaction logs Review past questions and answers to see how your bots are performing. ### Manage bot configurations View and retrieve detailed metadata and settings for every bot in your fleet. ### Search knowledge files Find specific knowledge sources by name keyword to quickly locate documentation. ## Use Cases ### Support Audit A support lead sees a weird answer and uses list_bot_interaction_logs to see the full history of that specific bot's recent queries to find the root cause. ### Content Verification A content manager wants to know if a new manual was added and uses list_bot_knowledge_sources to confirm the URL is active and included in the index. ### Bot Discovery A new hire wants to know what bots are available and asks the agent to list_team_bots to see the full catalog of available help resources. ### Knowledge Gap Analysis A product manager asks the agent to ask_bot_question about a new feature to see if the bot knows it yet, identifying where new docs are needed. ## Benefits - Stop manual checking: Use list_recently_indexed_bots to see what's live without refreshing your browser. This saves you from wondering if your latest documentation update actually went through. - Audit bot behavior: Pull real interaction logs with list_bot_interaction_logs to see how your customers are actually talking to your bots. It helps you spot where the bot might be getting confused. - Verify source data: See exactly what files are being used via list_bot_knowledge_sources to ensure your bot isn't hallucinating from old data. You get total visibility into the source material. - Quick bot testing: Use ask_bot_question to test new documentation updates immediately within your chat. You can verify answers in seconds before your team even sees the bot. - Scale your knowledge: Get a high-level view of your data size with get_bot_knowledge_summary to plan your content roadmap. It makes it easy to see where you need more documentation. ## How It Works The bottom line is you get a direct interface to manage your AI knowledge base without leaving your chat window. 1. Grab your API Key from the DocsBot account settings. 2. Connect the Connector to your AI client and authorize the connection. 3. Start asking your agent to manage your bots, check sources, or query your knowledge base. ## Frequently Asked Questions **Can DocsBot help me see if my documentation is actually being used?** Yes, you can use it to see exactly which URLs and PDFs are powering your bots. It gives you a clear view of your coverage. **How do I know if my bot just finished updating?** You can ask your agent to check the recently indexed bots to see the exact timestamp of the last update. This confirms your changes are live. **Can I use this to see what my customers are asking the bots?** Yes, you can pull the interaction logs to see a history of questions and answers. This is great for identifying common pain points. **Is this good for managing multiple different bots?** It's perfect for that. You can list all bots in a team or get details on any specific one, making it easy to oversee a large fleet. **Can I search for a specific document in my knowledge base?** Yes, you can search for sources by keyword to find specific files quickly. It saves you from scrolling through long lists of documentation. **Does this work for different types of files?** It tracks various sources, including URLs and PDFs, so you can see your full coverage across different types of documentation. **How do I get a DocsBot API Key?** Log in to your DocsBot account, navigate to the API section in your account settings, and you can generate or retrieve your unique API Key from there. **Can the agent train my bots?** This integration currently focuses on querying bots and auditing sources. Adding new sources or triggering full training cycles should be managed via the DocsBot dashboard. **Which AI models are used?** DocsBot supports various models including GPT-3.5 and GPT-4. The agent will retrieve results based on the specific model configured for your bot in your account.